Handbook of

Experimental Pharmacology

Volume 122

Editorial Board
G.V.R.Born,London
P. Cuatrecasas, Ann Arbor, MI
D. Ganten, Berlin
H. Herken, Berlin
K.L. Melmon, Stanford, CA
Springer
Berlin
Heidelberg
New York
Barcelona
Budapest
Hong Kong
London
Milan
Paris
Santa Clara
Singapore
Tokyo
Mechanisms
of Drug Interactions
Contributors
G.L. Amidon, H.H. Blume, J.R. erison, P.F. D'Arcy
P.A.G.M. De Smet, J.D. Ferrero, J.P. Griffin, e.M. Hughes
E. Lipka, J.e. McElnay, M. Schorderet, B.S. Schug, A. Somogyi
P.G. Welling, S. Yosselson-Superstine

Editors
P.F. D'Arcy, J.C. McElnay and P.G. Welling

Springer
ISBN-13: 978-3-642-64658-4 e-ISBN-13: 978-3-642-61015-8
DOI: 10.107/978-3-642-61015-8

softcover reprint of the hardcover 1st edition 1996

Preface

Over the years a number of excellent books have classified and detailed drug-
drug interactions into their respective categories, e.g. interactions at plasma
protein binding sites; those altering intestinal absorption or bioavailability;
those involving hepatic metabolising enzymes; those involving competition or
antagonism for receptor sites, and drug interactions modifying excretory
mechanisms. Such books have presented extensive tables of interactions and
their management. Although of considerable value to clinicians, such publica-
tions have not, however, been so expressive about the individual mechanisms
that underlie these interactions.
It is within this sphere of "mechanisms" that this present volume
specialises. It deals with mechanisms of in vitro and in vivo, drug-drug, drug-
food and drug-herbals interactions and those that cause drugs to interfere with
diagnostic laboratory tests. We believe that an explanation of the mechanisms
of such interactions will enable practitioners to understand more fully the
nature of the interactions and thus enable them to manage better their clinical
outcome.
If mechanisms of interactions are better understood, then it may be pos-
sible for the researcher to develop meaningful animal/biochemical/tissue cul-
ture or physicochemical models to which new molecules could be exposed
during their development stages. The present position, which largely relies on
patients experiencing adverse interactions before they can be established or
documented, can hardly be regarded as satisfactory.
This present volume is classified into two major parts; firstly, pharmacoki-
netic drug interactions and, secondly, pharmacodynamic drug interactions.
Within these parts, we have been fortunate to enlist the help of acknowledged
experts in preparing specific chapters focusing on aspects of the interaction
spectrum.
We believe that this volume will add much to that which is currently
known about drug interactions and will directly enhance the safer use of
medicines.

Belfast, Northern Ireland P.F. D'ARCY

Belfast, Northern Ireland J.e. McELNAY
Ann Arbor, MI, USA P.G.WELLING
List of Contributors

AMIDON, G.L., The University of Michigan, 2012 Pharmacy Bldg., Ann Arbor,
MI 48109-1065, USA

BLUME, H.H., Zentrallaboratorium Deutscher Apotheker, Ginnheimer StraBe

20, D-56760 Eschborn, Germany

CRISON, J.R., TSRL Inc., Ann Arbor, MI 48108, USA

D'ARCY, P.F., The Queen's University of Belfast, School of Pharmacy, Medi-

cal Biology Centre, 97 Lisburn Road, Belfast BT9 7BL, Northern Ireland

DE SMET, P.A.G.M., Drug Information Center, KLMP, Alexanderstraat 11,

NL-2514 JL The Hague, The Netherlands

FERRERO, J.D., Department of Pharmacology, University Medical Center, CH-

1211 Geneva 4, Switzerland

GRIFFIN, J.P., "Quartermans", Digswell Lane, Digswell, Welwyn Garden City,

Hertfordshire, AL7 1SP, UK

HUGHES, C.M., The Queen's University of Belfast, School of Pharmacy, Medi-

cal Biology Centre, 97 Lisburn Road, Belfast BT9 7BL, Northern Ireland

LIPKA, E., The University of Michigan, 2032 Pharmacy Bldg., Ann Arbor, MI
48109-1065, USA, and TSRL Inc., Ann Arbor, MI 48108, USA

McELNAY, J.C., The Queen's University of Belfast, School of Pharmacy,

Medical Biology Centre, 97 Lisburn Road, Belfast BT9 7BL, Northern
Ireland

SCHORDERET, M., Department of Pharmacology, University Medical Center,

CH-1211 Geneva 4, Switzerland

SCHUG, B.S., Zentrallaboratorium Deutscher Apotheker, Ginnheimer StraBe

20, D-56760 Eschborn, Germany
VIII List of Contributors

SOMOGYI, A., Department of Clinical and Experimental Pharmacology, Uni-

versity of Adelaide, Adelaide 5005, Australia

WELLING, P.G., PharmacokineticsiDrug Metabolism, Parke-Davis Pharma-

ceutical Research, 2800 Plymouth Road, Ann Arbor, MI 48105, USA

YOSSELSON-SUPERSTINE, S., Clinical Pharmacy Studies, School of Continuing

Medical Education, Sackler Faculty of Medicine, Tel Aviv University,
P.O.B. 39040, Tel Aviv 69978, Israel
Contents

CHAPTER 1
Introduction
P.F. D' ARCY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . .. .. 1
A. A Widening Problem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
I. Scope of This Present Volume. . . . . . . . . . . . . . . . . . . . . . . . . . . 2
B. Pharmacokinetic Drug Interactions. . . . . . . . . . . . . . . . . . . . . . . . . . . 2
I. Drug-Drug and Nutrient-Drug Interactions
at the Absorption Site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
II. Drug Interactions at Plasma- and Tissue-Binding Sites. . . . . . . 4
III. Drug Interactions and Drug-Metabolising Enzymes. . . . . . . . . 4
IV. Interactions Involving Renal Excretory Mechanisms. . . . . . . . 4
C. Pharmacodynamic Drug Interactions. . . . . . . . . . . . . . . . . . . . . . . . . . 5
I. Drug-Drug Interactions at the Receptor
and Other Active Sites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
II. Synergistic Interactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
III. Drug Interactions In Vitro. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
IV. Age and Genetic Factors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
V. Interference with Laboratory Testing. . . . . . . . . . . . . . . . . . . . . 7
VI. Herbal and Other Non-orthodox Medicines. . . . . . . . . . . . . . . . 7
D. Comment.................................................. 8
I. Drug Interactions:
Hazardous and Expensive Use of Resources? . . . . . . . . . . . . . . . 8
II. Sources of Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
III. The Literature on Drug Interactions. . . . . . . . . . . . . . . . . . . . . . 9
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
x Contents

Section I: Pharmacokinetic Drug Interactions

CHAPTER 2
Drng Interactions in the Gastrointestinal Tract
and Their Impact on Drug Absorption and Systemic Availability:
A Mechanistic Review
E. LIPKA, J.R. CRISON, B.S. SCHUG, H.H. BLUME and G.L. AMIDON
With 9 Figures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
A. Introduction............................................... 13
B. Physicochemical Interactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
I. Complexation with Metal Ions. . . . . . . . . . . . . . . . . . . . . . . . . . . 18
II. Binding to Resins. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
III. Complexation with Bile Salts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
IV. Nonspecific Adsorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
C. Interactions with Drugs That Influence GI Motility and pH . . . . . . 21
I. Alteration of Gastrointestinal Motility. . . . . . . . . . . . . . . . . . . . . 21
II. pH Alteration in the Gastrointestinal Tract. . . . . . . . . . . . . . . . . 26
D. Interactions Between Drugs
That Share the Same Absorption Mechanism. . . . . . . . . . . . . . . . . . . 27
I. Passive Absorption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
II. Carrier-Mediated Absorption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
E. Interactions as a Function of Intestinal Metabolism. . . . . . . . . . . . . 33
F. Altered Absorption as a Result
of Drug-Induced Mucosal Changes. . . . . . . . . . . . . . . . . . . . . . . . . . . 37
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

CHAPTER 3
Drug-Food Interactions Affecting Drug Absorption
P.G. WELLING. With 23 Figures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
A. Introduction............................................... 45
I. Food. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
II. Dosage Form. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
B. Influence of Food on the Gastrointestinal Tract. . . . . . . . . . . . . . . . 46
C. Direct Effect of Food on Drug Absorption. . . . . . . . . . . . . . . . . . . . 49
D. Interactions Causing Reduced Drug Absorption. . . . . . . . . . . . . . . . 51
E. Interactions Causing Delayed Drug Absorption. . . . . . . . . . . . . . . . . 63
F. Interactions Causing Increased Drug Absorption. . . . . . . . . . . . . . . 77
G. Interactions Causing Accelerated Drug Absorption. . . . . . . . . . . . . 95
H. Cases in Which Food Has No Effect on Drug Absorption....... 98
I. Conclusions................................................ 110
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Contents XI

CHAPTER 4
Drug Interactions at Plasma and Tissue Binding Sites
Joe. McELNAYo With 7 Figures 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 125
A. Introduction 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 125
Bo Proteins Involved in Drug Binding 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 125
e. Influence of Plasma and Tissue Binding on Drug Kinetics 0 0 0 0 0 0 0 0 127
Do Displacement of Drugs from Binding Sites 000000000000000000000 129
Eo Therapeutic Consequences
of Plasma Binding Displacement Drug-Drug Interactions 00000000 130
I. Rapid Intravenous Infusion of Displacing Agent 0 0 0 0 0 0 0 0 0 0 0 133
II. Parenteral Administration of Displaced Drug
Having High Extraction Ratio 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 134
III. Therapeutic Drug Monitoring and Drug Displacement
from Plasma Binding Sites 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 134
F. Therapeutic Consequences
of Tissue Binding Displacement Interactions 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 136
Go Displacement of Drugs from Binding Sites
by Endogenous Materials 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 138
Ho Disease States and Altered Plasma Protein Binding 0 0 0 0 0 0 0 0 0 0 0 0 0 139
I. Conclusions 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 145
References 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 145

CHAPTER 5
Drug Interactions and Drug-Metabolising Enzymes
PoFo D' ARCY 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 151
A. General Introduction 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 151
Bo Extrahepatic Microsomal Forms of Cytochrome P450 0 0 0 0 0 0 0 0 0 0 0 154
C. Genetic Polymorphism 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 156
Do Age and Disease and Cytochrome P450 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 000 157
Eo Clinical Importance of Enzyme Induction or Inhibition 0 0 0 0 0 0 0 0 0 0 158
I. Enzyme Inducers 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 158
II. Enzyme Inhibitors 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 160
Fo Conclusions 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 000000000000000 164
References 00000000000000000000000000000000000000000000000000 0 166

CHAPTER 6
Drug Interactions Involving Renal Excretory Mechanisms
A. SOMOGYI. With 5 Figures 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 173
Ao Introduction 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 173
B. Mechanisms of Renal Excretory Clearance 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 173
XII Contents

I. Anatomy and Physiology of the Kidney. . . . . . . . . . . . . . . . . . . 173

II. Glomerular Filtration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
III. Tubular Secretion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
1. Physiological Considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . 174
2. Cellular Mechanisms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
3. Pharmacokinetic Evidence for Tubular Secretion. . . . . . . . . 176
IV. Tubular Reabsorption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
C. Drug Interactions Involving Tubular Secretion. . . . . . . . . . . . . . . . . 177
I. Organic Anions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
1. Probenecid.......................................... 177
2. Methotrexate........................................ 184
II. Organic Cations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
1. Cimetidine.......................................... 188
2. Ranitidine........................................... 194
3. Famotidine.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
4. Trimethoprim.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
5. Amiodarone......................................... 197
6. Quinine/Quinidine................................... 197
III. Organic Neutral Drugs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
1. Digoxin............................................ 198
D. Tubular Reabsorption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
I. Proximal Tubule Site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
1. Lithium............................................ 201
II. Distal Tubule/Collecting Duct Site. . . . . . . . . . . . . . . . . . . . . . . . 206
1. Urine pH and Flow Rate. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

Section II: Pharmacodynamic Drug Interactions

CHAPTER 7
Drug-Drug Interactions at Receptors and Other Active Sites
M. SCHORDERET and J.D. FERRERO. With 5 Figures. . . . . . . . . . . . . . . . . . 215
A. Introduction............................................... 215
B. Mechanisms of Pharmacodynamic Interactions. . . . . . . . . . . . . . . . . 215
I. Transmitter Systems.................................... 215
1. Noradrenergic Synapse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
2. Dopaminergic Synapse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
3. Serotonergic Synapse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
4. Cholinergic Synapse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
5. GABAergic Synapse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
II. Ion Channels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
Contents XIII

1. Cardiac Ion Channels and Antiarrhythmic Drugs. . . . . . . . . 224

2. Potassium Channels
and Drug-Induced Torsades de Pointes. . . . . . . . . . . . . . . . . . 225
3. ATP-Sensitive Potassium Channels. . . . . . . . . . . . . . . . . . . . . 226
III. Hormonal Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
1. Adrenal Corticosteroids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
2. Glycaemic Regulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
IV. Homeostatic Regulations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
1. Renal Haemodynamics
and Drug-Induced Acute Renal Failure. . . . . . . . . . . . . . . . . 228
2. Prostaglandins, Natriuresis and Antihypertensive Therapy. 229
3. Potassium Homeostasis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230

CHAPTER 8
Synergistic Drug Interactions
J.P. GRIFFIN and P.F. D'ARCY.. . ...... . ..... . . . . ....... . ...... . . 235
A. Introduction: "Mithridatium"................................ 235
B. Early Use of Probenecid with Penicillin. . . . . . . . . . . . . . . . . . . . . . . 236
e. Diuretic Combinations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
D. Co-trimoxazole............................................ 237
E. Combination Treatment in Control of Epilepsy. . . . . . . . . . . . . . . . . 238
F. Antitubercular Drugs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
G. Anti-Leprosy Treatments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
H. Peptic Ulcer Therapy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
I. Non-Insulin-Dependent (Maturity Onset) Diabetes. . . . . . . . . . . . . 242
J. Cancer Chemotherapy. .. ...... . . ..... . ....... . . ..... . .. .... 245
K. Beneficial Interactions: A Philosophical Approach. . . . . . . . . . . . . . 246
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246

CHAPTER 9
In Vitro Drug Interactions
J.e. McELNAY and e.M. HUGHES. With 7 Figures. . . . . . . . . . . . . . . . . . 249
A. Introduction............................................... 249
B. Incompatibility Interactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
e. In Vitro Drug Interactions with Pharmaceutical Packaging
and Intravenous Administration Equipment. . . . . . . . . . . . . . . . . . . . 252
I. General Properties of Plastics ............................ 252
II. General Properties of Glass. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
III. Mechanisms of Interaction with Pharmaceutical Packaging. . . 254
1. Sorption............................................ 257
XIV Contents

2. Leaching............................................ 262
3. Permeation.......................................... 264
4. Polymer Modification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
IV. Drug Interactions with Contact Lenses. . . . . . . . . . . . . . . . . . . . . 265
D. In Vitro Drug Interactions in Therapeutic Drug Monitoring
and Drug Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
I. Cyclosporin............................................ 265
II. Chloroquine........................................... 266
E. In Vitro Drug Interactions as a Result of Formulation Changes. . . 266
I. Problems with Sustained-Release Formulations. . . . . . . . . . . . . 268
II. Drug Excipient Interactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
III. Formulation Effects on Rectal Bioavailability . . . . . . . . . . . . . . 272
F. Conclusions................................................ 274
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274

CHAPTER 10
Age and Genetic Factors in Drug Interactions
J.e. McELNAY and P.F. D'ARCY. With 2 Figures. . ........ . . . . ...... 279
A. Introduction............................................... 279
B. Age....................................................... 279
I. Pharmacokinetics in the Elderly. . . . . . . . . . . . . . . . . . . . . . . . . . 284
1. Absorption.......................................... 284
2. Distribution......................................... 284
3. Metabolism.......................................... 287
4. Excretion........................................... 288
II. Pharmacodynamics in the Elderly. . . . . . . . . . . . . . . . . . . . . . . . . 289
III. Inappropriate, Unnecessary and Interacting Medication. . . . . . 291
IV. Logistics: Age and Medicine Consumption. . . . . . . . . . . . . . . . . 293
V. Long-Term Treatments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
e. Genetic Factors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
D. Comment................................................. 300
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301

CHAPTER 11
Drugs Causing Interference with Laboratory Tests
S. YOSSELSON-SUPERSTINE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
A. Introduction............................................... 305
B. Mechanisms of Drug-Test Interactions. . . . . . . . . . . . . . . . . . . . . . . . 306
I. PharmacologicalInterferences............................ 306
1. Effect on Glucose Determination. . . . . . . . . . . . . . . . . . . . . . 306
2. Effect on Uric Acid Determination. . . . . . . . . . . . . . . . . . . . . 306
Contents XV

3. Intramuscular Injections
and Muscle Enzyme Determination. . . . . . . . . . . . . . . . . . . . . 307
4. Effect of Drug-Drug Interactions. . . . . . . . . . . . . . . . . . . . . . . 307
II. Methodological Interferences. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
1. Colorimetric Interferences. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
2. Effect on pH of the Assay Environment. . . . . . . . . . . . . . . . . 314
3. Interference with Chromatographic Methods. . . . . . . . . . . . . 315
4. Interference with Enzymatic Reactions. . . . . . . . . . . . . . . . . . 317
5. Interference with Immunoassays. . . . . . . . . . . . . . . . . . . . . . . 319
C. Design of a Study for Evaluation of Analytical Interference. . . . . . 320
D. Conclusions................................................ 321
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322

CHAPTER 12
Drug Interactions with Herbal and Other Non-orthodox Remedies
P.A.G.M. DE SMET and P.F. D'ARCY............................. 327
A. Introduction............................................... 327
B. Animal Agents. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
I. Fish Oil.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
II. Chinese Toad Venom. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
C. Amino Acids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
I. L-Tryptophan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
D. Vitamins.................................................. 330
E. Minerals.................................................. 331
F. Dietary Fads. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
G. Herbal Drugs....... . . . . . ..... . . . .... . . . ........ . ...... . . .. 332
I. Effects of Herbal Drugs
on Orthodox Drug Pharmacokinetics:
Effects on Absorption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
1. Dietary Fibres. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
2. Guar Gum......... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
3. Tannins............................................. 334
II. Effects of Herbal Drugs
on Orthodox Drug Pharmacokinetics:
Effects on Elimination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
1. Cola Nut. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
2. Eucalyptus Species. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
3. Grapefruit Juice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
4. Herbal Smoking Preparations. . . . . . . . . . . . . . . . . . . . . . . . . . 336
5. Kampo Medicines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
6. Liquorice............................................ 337
III. Effects of Orthodox Drugs
on Herbal Drug Pharmacokinetics. . . . . . . . . . . . . . . . . . . . . . . . 338
XVI Contents

1. Caffeine-Containing Herbs. . . . . . . . . . . . . . . . . . . . . . . . . . . . 338

2. Sparteine-Containing Herb. . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
3. Teucrium chamaedrys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
IV. Pharmacodynamic Interactions Between Herbal Drugs
and Orthodox Drugs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
1. Betel Nut (Areca catechu) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
2. Garlic (Allium sativum) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
3. Karela (Momordica charantia) . . . . . . . . . . . . . . . . . . . . . . . . . 341
4. Liquorice (Glycyrrhiza glabra) . . . . . . . . . . . . . . . . . . . . . . . . . 341
5. Picrorhiza kurroa... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
V. Multiple or Unclarified Interactions Between Herbal Drugs
and Orthodox Drugs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
1. Anthranoid Laxatives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
2. Berberine........................................... 342
3. Ginseng (Pamax ginseng) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
4. Piperine............................................ 342
5. "Shankhapusphi" . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
6. yohimbine.......................................... 343
VI. Interactions Between Different Herbal Drugs. . . . . . . . . . . . . . 343
H. Comment.................................................. 344
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346

Subject Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353

CHAPTER 1
Introduction
P.F. D'ARCY

A. A Widening Problem
Over 20 years ago, an editorial on drug interactions in the Lancet said that the
"publication of huge lists and tables will induce in doctors a drug-interaction-
anxiety syndrome and lead to therapeutic paralysis". This prediction has not
come about although the problem of drug interactions is still with us and the
spectrum is widening as new drugs are introduced. Indeed it could be said that
the nature of the problem has also widened for in the intervening years drug
interactions have come to embrace interactions with food and with herbal
medicines as well as the more numerous and better recognised drug-drug
interactions.
One of the advantages of the attention that has been focused on adverse
drug interactions over the past 20 years or so is that many drug-drug and drug-
food interactions are now predictable and many of the unwanted conse-
quences of using drug combinations can be avoided by simply adjusting the
dosage of one or more of the interactants. As a result of this, there has been a
considerable improvement in the safety and efficacy of therapy with drug
combinations.
Nowadays the focus on drug interactions tends to be more on their haz-
ards than their advantages. This is probably the correct orientation since the
vast majority of interactions are hazardous. An awareness of the possible
hazards of medication and possible interactions between drugs on the part of
those who use them, doctors and other health professionals, can only result in
better therapy with benefit to the patient in terms of both safety and efficacy.
Many excellent publications have explored the nature of drug interactions and
their message has filtered down to the patient level, being expressed by the
maxim "do not use two drugs when one will do".
Whereas over the years books have classified and detailed drug-drug
interactions into their respective categories, e.g., interactions at plasma pro-
tein-binding sites; interactions altering intestinal absorption or bioavailability;
interactions involving the hepatic-metabolising enzymes; interactions involv-
ing competition or antagonism for receptor sites, and finally drug interactions
modifying excretory mechanisms, they have not been so expressive about the
individual mechanisms that underlie these interactions. Their prime objective
was to inform about the interaction and its clinical management.
2 P.F. D'ARCY

I. Scope of This Present Volume

In their preliminary thoughts and discussions in the planning of this book, the
editors decided to attempt to produce a volume that dealt, specifically, with
the mechanisms underlying drug interactions. It was decided not to produce
extensive tables of interactions since this has already been competently done
by others. It was thought that an explanation of mechanisms would enable
practitioners, in some instances, to work out from first principles the nature
and management of interactions rather than to rely entirely on consulting long
tables of interactions and their management.
In recent years there has been intensive research into the mechanisms of
established drug-drug interactions in man. This present volume is classified
into two major parts: firstly, pharmacokinetic drug interactions and, secondly,
pharmacodynamic drug interactions.

B. Pharmacokinetic Drug Interactions

I. Drug-Drug and Nutrient-Drug Interactions at the Absorption Site

Chapter 2 discusses the many facets of drug-drug interactions at the absorp-
tion site. For the majority of drugs this is the stomach and small intestine. The
original intent of this book was to produce a single chapter on interactions
affecting drug absorption which would include drug-drug as well as drug-food
interactions. However, the quantity of material to be covered was too great
and, for convenience, drug-drug and drug-food interactions have been pre-
sented in two separate chapters. Of necessity a degree of overlap is unavoid-
able and it is convenient in this present context to deal with Chaps. 2 and 3
together.
The influence of wide-ranging factors such as the effect of fluid volume of
gastrointestinal physiology, splanchnic blood flow, passive diffusion, gastric
emptying time, pH of the intestinal contents, intestinal motility, ionic content
of foods, dietary fat, and gastrointestinal disease can have considerable effect
on the absorption of drugs and they are mechanisms common to both drug-
drug and drug-food interactions in the gut.
The two topics of drug-food interactions affecting drug absorption and the
effect of drugs on food and nutrient absorption are closely related but quite
distinct. In this present volume only the former topic is addressed (Chap. 3).
No attempt has been made to discuss the latter topic; material on this latter
topic has been reviewed elsewhere (BASU 1988; ROE 1989; TROVATO et al.
1991).
It is only in relatively recent years that studies on food-drug interactions
have been actively pursued. This work was heralded by the major review of
WELLING (1977). It is surprising that the study of drug-food interactions should
have lagged behind that of drug-drug interactions. It is noticeable that food-
drug interactions are not as often reported as drug-drug interactions although
Introduction 3

their potential frequency is far greater since food is by far the most common
substance associated with the ingestion of oral doses of medicine. In my
experience, medical and pharmacy practitioners are not well informed about
food-drug interactions and therefore may fail to pass on cautionary advice to
their patients.
Perhaps one of the most troublesome and well reported of food-drug
interactions was the interaction between tyramine-containing foods and
monoamine oxidase inhibitor (MAOI) antidepressants during the early 1960s
(see reviews by D' ARCY 1979; BROWN et al. 1989; LIPPMAN and NASH 1990).
Deaths from eating cheese whilst taking MAOIs was given great publicity and
this type of interaction did much to create an awareness of the potential of
foods to interact with medicines.
Monoamine oxidase type A preferentially oxidatively de-aminates
adrenaline, noradrenaline and serotonin, while type B preferentially
metabolises benzylamine and phenylethylamine. Dopamine and tyramine are
de-aminated by both forms of the enzyme. The traditional MAOIs, such as
phenelzine and tranylcypromine, inhibit both types A and B of the enzyme.
Selective MAOIs, such as brofaromine, moclobemide and clorgyline, have
been introduced with a selective and reversible inhibition of monoamine
oxidase (type A), and selegiline, which reversibly inhibits type B, is an adjunct
in the treatment of parkinsonism. The dietary restrictions that need to be
observed with the traditional MAOIs (type A and B) are much less stringent
with the new selective inhibitors.
It was largely the interactions with food that led to a virtual replacement
of the use of MAOI antidepressents by the tri- and polycyclic antidepressants.
That this may have been a collective error of judgement is now being realised;
perhaps the MAOIs are not so bad after all!
The effect of heavy metal ions in milk and other dairy products on the
absorption of tetracyclines was another milestone which raised the sensibility
of clinicians to the dangers of food-drug interactions (BRAYBROOKS et al. 1975;
CHIN and LACH 1975; NEUVONEN 1976).
I well recall in the late 1940s, after tetracycline had been first introduced,
advising mothers to open the capsules of tetracycline and dissolve the contents
in a little milk before giving the antibiotic to their infants. But it took almost
20 years before the milk-tetracycline interaction was recognised. Once this
interaction was realised, it was only a relatively short time before the spectrum
of such interactions was widened and it was realised that it was not uncommon
that the absorption from the gut of other antimicrobial agents, or reduc-
tion of their bioavailability, could be seriously influenced by food or metal
ions. Even more recent work has shown that the bisphosphonates, used to
treat a broad range of bone disorders, including osteolytic bone metastases,
hyperparathyroidism, Paget's disease of bone and established vertebral
osteoporosis in women, can be completely antagonised by food and vitamin-
mineral supplements with a high calcium content (FOGELMAN et al. 1984, 1985,
1986; FELS et al. 1989; COMPSTON 1994).
4 P.F. D'ARCY

II. Drug Interactions at Plasma- and Tissue-Binding Sites

Drug interactions at plasma- and tissue-binding sites and their mechanisms are
presented in Chap. 4. The importance of plasma protein-binding displacement
as a clinically important drug interaction mechanism is controversial.
When a displacing agent interacts with a primary drug the result is an
increase in the free concentration of the displaced drug in the plasma. In the
past, it has been wrongly assumed that these free drug concentrations will
remain raised. This is not so. What in fact occurs is that the increased free drug
in the plasma quickly distributes throughout the body and may localise in the
tissues. When equilibrium is again reached C( assuming the volume into which
the drug distributes is large) the free concentration in the plasma will be the
same, or near to its pre-interaction level. The "total" drug concentration (i.e.
bound plus free drug) in the plasma will decrease after the equilibrium for free
drug is reached. The ultimate consequence, therefore, of a plasma protein-
binding displacement interaction is to lower total drug concentration in the
plasma and to leave free drug concentration unchanged. Since pharmacologi-
cal actions, including toxicities, correlate with free drug in the plasma, any
increased effect seen after displacement is usually of a transient nature and, in
most cases, is clinically unimportant.
Generally, therefore, although current opinion accepts that the binding of
drugs to protein is an important pharmacokinetic parameter, it is of the view
that its importance as a mechanism of drug-drug interaction has been over-
estimated and overstated.

III. Drug Interactions and Drug-Metabolising Enzymes

It is now realised that many interactions can be explained by alterations in the
metabolic enzymes that are present in the liver and other extra-hepatic tissues.
Induction or inhibition of the collection of isoenzymes, collectively known as
cytochrome P450 enzymes, are mechanisms that have been shown to underlie
some of the more serious drug-drug interactions.
Chapter 5 presents a chronological account of research in this area, and
discusses extrahepatic microsomal forms of cytochrome P450, genetic poly-
morphism, age and disease and cytochrome P450, and the clinical importance
of enzyme induction or inhibition. Although the chapter concentrates on
the P450 enzyme systems, it emphasises that certain oxidative reactions of
xenobiotics are also catalysed by enzymes other than cytochrome P450s. A
brief account of these other enzymes is also presented.

IV. Interactions Involving Renal Excretory Mechanisms

The kidney represents the final elimination organ for virtually all foreign
substances irrespective of whether they are cleared unchanged or as metabo-
lites formed in the body predominantly by the liver. Chapter 6 presents an
Introduction 5

account of the mechanisms of renal excretory clearance, drug interactions

involving tubular secretion and tubular reabsorption.
Such interactions can have important clinical implications in terms of
patient mortality and morbidity. It is fortunate that an understanding of the
molecular mechanisms of renal drug transport has allowed predictions to
be made of potential drug interactions in early phase drug development. This
understanding together with the availability of in vitro models of renal drug
transport has also been used as an early guide to potential in vivo drug
interactions. This chapter has also highlighted inconsistencies in preconceived
knowledge on the renal tubular secretion of drugs.

C. Pharmacodynamic Drug Interactions

I. Drug-Drug Interactions at the Receptor and Other Active Sites

Pharmacodynamic interactions at receptor sites have been extensively re-
viewed in Chap. 7. Although the majority of drug-drug interactions are poten-
tially, if not actually, hazardous, some are synergistic and may be to the
patient's therapeutic advantage. The chapter concentrates on adverse drug
interactions resulting in either a diminished therapeutic effect or an increased
toxicity. The chapter describes the mechanisms by which pharmacodynamic
interactions occur and illustrates them with clinically relevant interactions.
Theoretically, at least, since interactions are based on supposedly established
mechanisms, they should be more easily prevented and should have less
clinical impact than pharmacokinetic interactions. However, the relevance of
such a view may be questioned in the light of recent advances in the
characterisation and detection of kinetic interactions.

II. Synergistic Interactions

Not all drug-drug interactions are hazardous, and some synergistic inter-
actions when they occur can be of clinical value in therapy. Indeed some
food-drug interactions can be valuable especially when food improves the
bioavailability of the active drug substance. In other examples, the inhibition
of the oxidative phase of hepatic drug metabolism by, for example, cimetidine
may potentiate the effect and/or duration of a variety of drugs. Unfortunately,
increasing the plasma concentration of a primary drug may well also increase
the probability of enhanced toxicity.
In current medical practice there are a number of examples of combined
formulations and/or co-prescribing of active ingredients on the basis of
their believed synergistic action. For many of these the evidence for efficacy
may not stand up to critical analysis. A brief account of the more important
treatments with believed synergistic drug combinations is presented in
Chap. 8.
6 P.F. D'ARCY

III. Drug Interactions In Vitro

Currently much information is appearing in the literature usually from phar-
maceutical research sources on drug interactions in vitro. Chapter 9 deals with
this and will update the reader on the mechanisms involved in the latest
developments. Much of this information in the past has been interactions
between drug and drug, and drug and fluid in intravenous infusions. Newer
work has concentrated on interactions between specific drugs, notably
chloroquine, cyclosporin insulin and vasodilator nitrates, and pharmaceutical
packaging materials (glass and plastics), and the mechanisms by which these in
vitro interactions occur.
Also included in the chapter are drug interactions with contact lens mate-
rials and drug-excipient interactions. In the past, it has not always been appre-
ciated that the so-called "inactive" excipients used in product formulations are
physicochemically active substances which are often quite capable of interact-
ing or producing complexes with drug substances in the formulation. They can
cause changed absorption and altered bioavailability. The often quoted out-
break of phenytoin intoxication in Australia during 1968-1969 was caused by
increased bioavailability due to a simple change in capsule filler from calcium
sulphate to lactose (TYRER et al. 1970).
For some time now, a great deal of effort has been given to the complexing
of drug substances to a variety of polymers with the object of producing
sustained-release products. Some of this research has been done with proteins,
the ultimate goal, I believe, being the production of an orally effective insulin.

IV. Age and Genetic Factors

Particular attention in this volume has been given to the influences of age and
genetic factors in drug interaction (Chap. 10). It has long been established that
elderly patients use more medicines than younger age groups. It is, however,
clear from this chapter that age and genetic factors are not directly responsible
for causing drug-drug interactions and that, qualitatively, the elderly share
similar drug interactions with younger age groups. It is the severity of the
interaction that may be more serious in the elderly patient or in the patient
with a genetic abnormality. Also the increased use of medication by the elderly
allows a greater possibility of drug-drug interactions occurring.
It may be noted that most of the drugs involved in clinically relevant
interactions are those upon which patients are carefully stabilised for rela-
tively long periods of time, for example, anticoagulants, anticonvulsants, beta-
blockers and other hypotensives, cardiac glycosides, Hz-receptor blockers,
hypoglycaemics, lithium salts, psychotropics and immunosuppressants. It may
be deduced that many of these patients will be in the older age bracket.
Examination of clinical reports of drug interactions often shows that it is these
drug-stabilised patients who are at special risk of any changes in therapy or
Introduction 7

environment that will influence the potency or availability of their normal

medication. It must also be appreciated that removal of a drug from an
otherwise stabilised regimen of treatment may also initiate a serious inter-
action sequel.

V. Interference with Laboratory Testing

Laboratory tests, physical examination and the patient's medical history often
provide the main key to accurate diagnosis and to subsequent rational therapy.
Chapter 11 relates how one of these key features - laboratory tests - can be
influenced in sensitivity or specificity (or both) by the drugs consumed by the
patient. Frequently the influence of medication on the reliability of subsequent
laboratory tests is not realised or merely overlooked. As the chapter details,
the mechanisms of such interactions can be generally classified into two areas:
pharmacological and methodological interferences. The author of this chapter
has emphasised that whenever a drug-laboratory test interaction is established
it should be communicated to the medical community so that they may avoid
possible errors in diagnosis. Chapter 11 summarises much information about a
field of interaction, the importance of which is little realised or appreciated.

VI. Herbal and Other Non-orthodox Medicines

There are many types of herbal remedies. At one end of the scale there are the
self-made teas prepared from self-collected herbs. At the other end there are
officially registered drug products which have passed through a rigid registra-
tion procedure. In between there are a bewildering range of nostrums that
have obscure origins and even more obscure constituents. The latter classes
may be a mixture of herbal products with other ingredients of non-herbal
origin (e.g. arsenic or lead) or undeclared Western drugs [e.g., corticoster-
oids, non-steroidal anti-inflammatory/antirheumatic agents (NSAIDS) or
paracetamol]. Although herbal medicines are by far the largest component
of non-orthodox remedies, they do not have exclusive claims. Other non-
orthodox remedies range from preparations of animal origin, minerals, vita-
mins and amino acids; many of these are capable of interfering with orthodox
medicines.
A survey by WHARTON and LEWITH (1986) showed that only 5% of British
doctors claimed more than a poor knowledge of herbal medicines. Medical
practitioners generally believed that herbal preparations were residues from
the remote past, that they were harmless and largely ineffective. Many patients
believed that they were safe, simply because they were natural products, and
that they were often miraculously effective. Unfortunately both patients and
doctors are misinformed since, as Chap. 12 relates, many herbal products can
be exceedingly toxic and may indeed present a peculiar hazard if taken in
combination with orthodox medicines.
8 P.F. D'ARCY

D. Comment

I. Drug Interactions: Hazardous and Expensive Use of Resources?

There is little doubt that drug-drug interactions can often be serious even life-
threatening. They can also be very expensive and evidence from the Medical
Defence Union's Reports of over 10 years ago reveals that one case which
settled for £44000 was due to phenylbutazone-induced potentiation of war-
farin which was followed by an intraspinal haemorrhage, resulting in an in-
complete tetraplegia at the level of C7. It is interesting in this case that the
solicitor's letter stated that "Butazolidin is a well known potentiator of
counmarin anticoagulants of which warfarin is one" ... "the prescribing of
Butazolidin for a patient known to be taking warfarin routinely was a breach
of your professional duty to him".
Knowledge of the "state of the art" has much advanced since the 1980s
and what were then novel and scientifically interesting interactions are now
established and well-recorded interactions in standard books of reference.
Knowledge has advanced but their has been little progress in the way in which
it is generated.

II. Sources of Information

Most reports of drug-drug or drug-food interaction arise primarily from ex-
perience in the clinic. Some of these are anecdotal and may not be reported
again, whilst others represent the first warning signs of a wider cohort of cases
that is yet to be realised. Their purpose is to report the interaction and explain
its hazard so that others can avoid it occurring. However, few of these reports
are able to indicate the precise mechanism of the interaction.
Generally, animal experiments have not contributed much of substance to
knowledge about interactions. Admittedly, the early work of CONNEY et al.
(1956, 1957) in rats did much to explore the nature and organisation of enzyme
systems in liver microsomes, but it was the human studies of LEVY (1970) and
NEUVONEN et al. (1970) that provided the now classical evidence that the salts
of divalent or trivalent metals formed non-absorbable complexes with tetracy-
clines and reduced their absorption, resulting in subtherapeutic levels in the
plasma. It was observation in patients that gave the first indication that
concomitant medication could antagonise the efficacy of oral contraceptives
(DOSSETOR 1975). It was in tuberculosis patients that the apparent interaction
between para-aminosalicylic acid (PAS) and rifampicin was first recognised
and later to be explained that the interaction was not due to PAS but to the
bentonite content of its granules (BOMAN et al. 1971). I was reminded of this
latter interaction in recent months when I met a lady in Belfast who had been
a TB patient at that time and had been receiving rifampicin and PAS and
without warning or explanation was changed to rifampicin plus isoniazid.
Introduction 9

III. The Literature on Drug Interactions

The literature on drug interactions tends to be very non-specific and in the last
20 years or so it has become voluminous and has been cluttered up with a
sticky mass of largely irrelevant studies of interactions in animal models and in
single-dose pharmacokinetic studies in animals and in young adult age groups
of volunteers. Such studies can be of predictive value, but only if they mimic
the clinical situation and if they relate to drug combinations and dosage
regimens that are normally used in sick patients.
It is clear that the literature requires some judgmental element to be
exercised as to its clinical significance. It is the single reports in the correspon-
dence columns of medical and pharmaceutical journals that, although often
anecdotal and uncorroborated, have, in my view, made a major contribution to
elucidating the nature and extent of interactions. Such reports are also useful
since they often stimulate other clinicians to report similar experiences in their
own patients.
There is still need to focus attention on those drug-drug or drug-nutrient
interactions that really do influence the safety or efficacy of human drug
therapy in all age groups. If mechanisms of interaction are clearly understood
then it may be possible to develop animal/biochemical/tissue culture/physico-
chemical or other types of models of interactions to which new molcules could
be exposed during their development stages. It is hardly satisfactory or safe to
have to rely on patients experiencing adverse interactions before they can be
established and documented. It is hoped that, by discussing mechanisms of
interactions, this volume will stimulate interest in investigating the possibility
and potential of developing meaningful models.

References
Basu TV (1988) Drug nutrient interactions. Croom Helm, London
Boman G, Hanngren A, Malmborg A, Borga 0, Sjoqvist F (1971) Drug interaction:
decreased serum concentrations of rifampicin when given with PAS. Lancet
1:800
Braybrooks MP, Barry BW, Abbs ET (1975) The effect of mucin on the bioavailability
of tetracycline from the gastrointestinal tract: in vivo, in vitro correlation. J Pharm
Pharmacol 27:508---515
Brown C, Taniguchi G, Yip K (1989) The monoamine oxidase inhibitor-tyramine
interaction. J Clin Pharmacol 29:529-532
Chin TF, Lach JL (1975) Drug diffusion and availability: tetracycline metallic chela-
tion. Am J Hosp Pharm 32:625-629
Compston JE (1994) The therapeutic use of bisphosphonates. Br Med J 309:711-715
Conney AH, Miller EC, Miller JA (1956) The metabolism of methylated aminoazo
dyes. V. Evidence for induction of enzyme synthesis in the rat by 3-methylcholan-
threne. Cancer Res 16:450-459
Conney AH, Miller EC, Miller JA (1957) Substrate-induced synthesis and other
properties of benzpyrenehydroxylase in rat liver. J BioI Chern 228:753-766
D'Arcy PF (1979) Drug interactions. In: D'Arcy PF, Griffin JP (eds) Iatrogenic dis-
eases, 2nd edn. Oxford University Press, Oxford, pp 45-76
Dossetor J (1975) Drug interactions with oral contraceptives. Br Med J 4:467-468
10 P.F. D'ARCY: Introduction

Fels JP, Necciari J, Toussain P, Debry G, Luckx A, Scheen A (1989) Effect of food
intake on kinetics and bioavailability of (4-chlorophenyl) thiomethylene
bisphosphonic acid (Abstr). Calcif Tissue Int 44 [Suppl]:S-104
Fogelman I, Smith ML, Mazess R, Bevan JA (1984) Absorption of diphosphonate in
normal subjects (Abstr). Calcif Tissue Int 36 [SuppI2]:574
Fogelman I, Smith ML, Mazess R, Bevan JA (1985) Absorption of oral diphosphonate
in normal subjects (Abstr). Bone 6:54
Fogelman I, Smith ML, Mazess R, Wilson MA, Bevan JA (1986) Absorption of oral
diphosphonate in normal subjects. Clin Endocrinol (Oxf) 24:57-62
Levy G (1970) Biopharmaceutical considerations in dosage form and design. In:
Sprowls JB (ed) Prescription pharmacy, 2nd edn. Lippincott, Philadelphia, pp 70,
75, 80
Lippman SB, Nash K (1990) Monoamine oxidase inhibitor update: potential adverse
food and drug interactions. Drug Saf 5:195-204
Neuvonen PJ (1976) Interactions with the absorption of tetracyclines. Drugs 11:45-54
Neuvonen PJ, Gothoni G, Hackman R, Bj6rksten K (1970) Interference of iron with
the absorption of tetracyclines in man. Br Med J 4:532-534
Roe AR (1989) Diet and drug interactions. Van Nostrand Reinhold, New York
Trovato A, Nuhlicek DN, Midtling JE (1991) Drug-nutrient interactions. Am Fam
Physician 44:1651-1658
Tyrer JH, Eadie MJ, Sutherland JM, Hooper WD (1970) Outbreak of anticonvulsant
intoxication in an Australian city. Br Med J 2:271-273
Welling PG (1977) Influence of food and diet on gastrointestinal drug absorption. A
review. J Pharmacokin et Biopharm 5:291-334
Wharton R, Lewith G (1986) Complementary medicine and the general practitioner.
Br Med J 292:1498-1500
Section I
Pharmacokinetic Drug Interactions
CHAPTER 2
Drug Interactions in the Gastrointestinal Tract
and Their Impact on Drug Absorption and
Systemic Availability: A Mechanistic Review
E. LIpKA, l.R. CRISON, B.S. SCHUG, H.H. BLUME, and G.L. AMIDON

A. Introduction
The effect of drug interactions with other drugs, excipients and gastrointesti-
nal contents on their oral absorption and systemic availability have been
extensively investigated over the past 50 years. These interactions can be
classified as direct interactions between the drug and other drugs or compo-
nents in the formulation, e.g., excipients in the gastrointestinal tract altering
the drugs' thermodynamic activity for absorption or, alternatively, the effects
may be indirect through alteration of gastrointestinal transit, gastrointestinal
secretions, or gastrointestinallhepatic metabolism or elimination. Recently
attention has been directed to metabolism interactions in the liver and even
more recently in the gastrointestinal mucosal tissue. These interactions can
be caused by direct inhibition of enzyme metabolism or enzyme induction,
resulting in changes in the absorption rate and leading to altered oral
bioavailability. These metabolic effects also can be indirect, through the
position dependence of metabolizing enzyme levels in the gastrointestinal
tract; consequently transit changes can alter metabolic profiles and systemic
availability. Recent interest has been directed to carrier-mediated absorption,
where drug interactions can occur through competition for carrieres) respon-
sible for absorption and, finally, it has been proposed that the P-glycoprotein
exporter may be responsible for some of the observed drug interactions
and nonlinear absorption effects. This review will focus on examples of the
various types of mechanistic interactions noted above. The interaction of
drugs with foods, a very important component in drug regulation today,
is reviewed elsewhere in this volume. A more comprehensive survey of
relevant drug-drug interactions than can be covered in this review can be
found in Table 1.

B. Physicochemical Interactions
The two physicochemical interactions that affect oral absorption arise due to
complexation of the drug with opposite-charged ion species or to nonspecific
adsorption of the drug. In this section, examples of complexation of drugs with
Table 1. Summary of drug-drug interactions influencing absorption. (Modified from WELLING 1984)

Drug affected Interfering agent Observed change in Possible cause Reference

absorption

p-Aminosalicylate (PAS) Diphenhydramine Delayed Delayed stomach LAVIGNE and MARCHAND (1973)
emptying
Aspirin Charcoal Reduced Adsorption NEUVONEN et al. (1978); LEVY and
TSUCHIYA (1972)
Aspirin (enteric-coated) Antacids Increased rate Faster drug release FELDMAN and CARLSTEDT (1974)
Carbamazepine Charcoal Reduced Adsorption NEUVONEN and ELONEN (1980)
Cefdinir Iron ion Reduced rate and Chelation VENO et al. (1993)
extent
Cephalexin Cholestyramine Reduced Adsorption or PARSONS and PADDOCK (1975)
steatorrhea
Chlorothiazide Colestipol Reduced Binding KAUFFMAN and AZARNOFF (1973)
Metoclopramide Decreased Absorption window OSMAN and WELLING (1983)
or dissolution
Propantheline Increased Absorption window OSMAN and WELLING (1983)
or dissolution
Chlorpromazine Antacids Reduced Adsorption FORREST et al. (1970); FANN et al.
Benzhexol Reduced Delayed stomach (1973) RIVERA-CALIMLIM et al.
(trihexphenidyl) emptying (1973)
Chlortetracycline Antacids Reduced Adsorption SEED and WILSON (1950);
GREENSPAN et al. (1951);
WAISBREN and HUECKEL (1950)
Ciprofloxacin Antacids Reduced Chelation LODE (1988) and NIX et al. (1989)
Clindamycin Kaolin-pectin Delayed Adsorption ALBERT et al. (1978b)
Diazepam Antacids Delayed Adsorption GREENBLATT et al. (1978)
Dicoumarol Magnesium hydroxide Increased Chelation AMBRE and FISCHER (1973)
(bishydroxycoumarin)
Digitoxin Cholestyramine Increased elimination Interrupted enterophepatic CALDWELL et al. (1971)
rate circulation
Digoxin Neomycin Reduced Sprue-like syndrome, LINDENBAUM et a1. (1972)
malabsorption
Antacids Reduced Adsorption and BROWN and JUHL (1976);
faster stomach ALBERT et a1. (1978a)
emptying (partly)
Charcoal Reduced Adsorption NEUYONEN et a1. (1978)
Metoclopramide Reduced Limited dissolution MANNINEN et a1. (1973)
Propantheline Increased Dissolution MANNINEN et a1. (1973)
Doxycycline Ferrous sulfate Reduced Chelation, influencing NEUYONEN and PENTTILA (1974);
absorption and NEUVONEN et a1. (1970)
elimination
Enoxacin Ranitidine Reduced Increased pH LEBSACK et a1. (1992)
Ethanol Metoclopramide Increased rate Faster stomach emptying GIBBONS and LANT (1975)
Propantheline Reduced rate Delayed stomach emptying GIBBONS and LANT (1975)
Atropine Reduced rate Delayed stomach emptying GIBBONS and LANT (1975)
Hydrochlorothiazide Propantheline Delayed, but Delayed stomach emptying BEERMAN and GROSCHINSKY-
increased GRIND (1978)
Ferrous ion Tetracycline Reduced Chelation NEUVONEN et a1. (1975); HEINRICH
et a1. (1974)
Isoniazid Antacids Delayed and reduced Adsorption and first-pass HURWITZ and SCHLOZMAN (1974)
metabolism
Levodopa Homatropine Reduced patient Delayed stomach emptying, FERMAGLICH and O'DOHERTY
response increased metabolism in (1972)
stomach
Antacids Increased Faster stomach emptying REVERA-CALIMLIM et a1. (1971)
Metoclopramide Increased rate Faster stomach emptying MORRIS et a1. (1976)
Lignocaine (lidocaine) General anesthetics Delayed Delayed stomach emptying ADJEPON-YAMOAH et a1. (1973)
Lincomycin Sodium cyclamate Reduced WAGNER (1969)
Lithium Metoclopramide Increased rate Faster stomach emptying CRAMMER et a1. (1974)
Propantheline Delayed Delayed stomach emptying CRAMMER et a1. (1974)
Methacycline Ferrous sulfate Reduced Chelation NEUVONEN et a1. (1970)
Nadolol Trihydroxy bile saits, Reduced Poorly soluble micelle YAMAGUCHI (1986)
sodium cholate formation
Table 1. Continued

Drug affected Interfering agent Observed change in Possible cause Reference

absorption
Nitrofurantoin Magnesium trisilicate Reduced Adsorption NAGGAR and KHALIL (1979)
Norfloxacin Antacids Reduced Chelation NIX et al. (1990)
Ofloxacin Antacids Reduced Chelation LODE (1988)
Oxytetracycline Ferrous sulfate Reduced Chelation NEuvONEN et al. (1970)
Paracetamol Metoclopramide Increased rate Faster stomach emptying NIMMO et al. (1973)
(acetaminophen) Propantheline Delayed Delayed stomach emptying NIMMO et al. (1973)
Pethidine Delayed Delayed stomach emptyin NIMMO et al. (1973)
Morphine Delayed Delayed stomach emptying NIMMO et al. (1973)
Penicillamine Ferrous sulfate Reduced Chelation SCHUNA et al. (1983)
Antacids Reduced Adsorption, chelation SCHUNA et al. (1983)
Penicillin V Neomycin Reduced Sprue-like syndrome, CHENG and WHITE (1962)
malabsorption
Phenobarbitone Charcoal Reduced absorption Adsorption (reduces NEuvONEN and ELONEN (1980)
and increased availability and prevents
elimination rate intestinal reabsorption)
Phenylbutazone Charcoal Reduced absorption Adsorption (reduces NEuvONEN and ELONEN (1980)
and increased availability and prevents
elimination rate intestinal reabsorption)
Phenytoin Charcoal Reduced
Pseudoephedrine Aluminum hydroxide Increased rate Raised GI pH LUCAROTII et al. (1972)
Kaolin Delayed Adsorption
Promazine Attapulgite and pectin Reduced Adsorption SORBY and Lm (1966)
Proquazone Antacids Delayed OHNHAUS (1980)
Quinine Antacids Reduced (in rats) Raised pH, delayed stomach HURWITZ (1971)
emptying
Riboflavine-5' - Sodium alginate Increased Slow GI transit because of LEVY and RAO (1972)
phosphate high viscosity
Rifampicin p-Aminosalicylate Delayed and reduced Adsorption to bentonite in BOMAN et al. (1971, 1975)
(PAS) PAS granules
Sulfadiazine Sodium bicarbonate Increased rate Faster dissolution rate PETERSON and FINLAND (1942)
Magnesium hydroxide Increase (in rats) Faster dissolution rate HURWITZ (1971)
Sulfadiazine sodium Antacids Decreased (in rats) Delayed stomach emptying HURWITZ (1971)
Sulfamethoxazole Cholestyramine Reduced Adsorption or steatorrhea PARSONS and PADDOCK (1975)
Sulfathiazole Antacids Increased Faster dissolution rate BARLOW and CLIMENKO (1941)
Tetracycline Ferrous sulfate Reduced Chelation NEUVONEN et al. (1970); GOTHONI
et al. (1972)
Ferrous salts Variable reduction Chelation NEUVONEN and TURAKKA (1974)
(depending on salt)
Zinc sulfate Reduced Chelation PENTTILA et al. (1975)
Magnesium sulfate Reduced HARCOURT and HAMBURGER
(1957)
Proteolytic enzymes Unaffected or BRADBROOK et al. (1978); SENECA
increased and PEER (1965)
Fe 3+, Fe2+, AP+, Reduced Chelation NEUVONEN (1976)
Mg2+, Ca2+
Sodium bicarbonate Reduced Poor dissolution BARR et al. (1971)
Thyroxine Cholestyramine Reduced Adsorption NORTHCUTT et al. (1969)
Trimethoprim Cholestyramine Delayed Adsorption PARSONS and PADDOCK (1975)
Vitamin A Neomycin Reduced Sprue-like syndrome, BARROWMAN et al. (1973)
malabsorption
Aluminum hydroxide Reduced Absorption oxidation HOFFMAN and DYNIEWICZ (1945)
Vitamin B12 Colchicine Reduced Ileal blockade FALOON and CHODOS (1969)
Neomycin Reduced Malabsorption FALOON and CHODOS (1969)
p-Aminosalicylate Reduced PALVA et al. (1972)
Warfarin Magnesium hydroxide No effect AMBRE and FISCHER (1973)
Cholestyramine Reduced Adsorption ROBINSON et al. (1971)
18 E. LIPKA et al.

metal ions, resins and bile salts as well as adsorption of drugs on charcoal will
be discussed.

I. Complexation with Metal Ions

Most tetracycline derivatives are rapidly, but incompletely, absorbed and the
extent of absorption is further reduced when the drug is co administered with
other medications or nutrients that contain bi- or trivalent cations, such as
calcium, magnesium and iron (DITCHBURN and PRITCHARD 1956). These cations
form complexes with tetracyclines that are either poorly absorbed and/or
poorly soluble, leading in some cases to clinically relevant reduction of plasma
levels up to 90% (WELLING 1984). The extent of absorption inhibition depends
upon the solubility and stability of the drug-metal complexes formed, which
in turn are determined by the physicochemical characteristics of both the
tetracyclines and the metal ions. For tetracycline, the order of decreasing
complex stability has been reported to be Fe3+> AP+ > Cu2+> Ni2+ > Fe2+> C02+
> Zn2+ > Mn 2+ (NEUVONEN 1976). Additionally, the stability of the metal salt
administered, which defines the free fraction of the cation, also plays a role
with regard to drug-metal complex formation. In a series of iron salts, ferrous
sulfonate has been shown to reduce tetracycline plasma levels by up to 90%,
whereas ferric sodium edetate induced a reduction of only 30% (NEUVONEN
and TURAKKA 1974). Tetracycline derivatives contain multiple potential bind-
ing sites for chelate formation, the most important one being the 1,3-keto-enol

Parameters Study 1 (n = 6) Study 2 (n = 6) Study 3 (n = 6)

(ug/ml)
e max 1.71 ± 0.42 0.16 ± 0.14 1.28 ± 0.23
(h)
{max 4.2 ± 1.0 1.8 ± 2.1 3.3 ± 0.5
AUCO-12 (ugh/ml) 10.3 ± 1.35*'** 0.78 ± 0.25 6.55 ± 1.61
AUC0-3 (jigh/ml) 2.13 ± 0.83*** 0.38 ± 0.3 1.95 ± 0.24
AUC3-12 (ugh/ml) 8.03 ± 1.72*,** 0.4 ± 0.31 4.60 ± 1.54

Data are mean values ± SD, emaX" maximum plasma concentration; {maX' time to reach
e max ; AUC, area under the plasma concentration-time curve. Study 1, cefdinir alone;
study 2, cefdinir and iron ion preparation (simultaneous administration); study 3, iron
preparation 3 h after cefdinir administration.
* p < 0,01, study 1 versus study 2; ** P < 0.05, study 1 versus study 3; *** P < 0,05, study
1 versus study 2.
Fig. 1. Structures of cefdinir and the formation of cefdinir-iron ion complex. The table
summarizes the pharmacokinetic parameters for cefdinir. (Adapted from UENO et al.
1993)
Drug Interactions in the Gastrointestinal Tract 19

structure (WEINBERG 1957). Drug-metal complexes of 1: 1 and at higher pH of

1 : 2 might be formed. With trivalent cations tetracycline forms complexes of
1: 3. Some of these complexes are poorly soluble in water, while others, al-
though moderately soluble, exhibit a reduced diffusion rate as a consequence
of complexation (CHIN and LACH 1975).
Cefdinir, a new oral cephalosporin, has been shown to undergo a 93 %
reduction of bioavailability when co administered with iron ions (VENO et al.
1993). Since the antibiotic is administered as a sustained-release dosage form,
it reaches maximum plasma levels after 4h, and even administration of iron
preparations 3h after cefdinir intake reduces bioavailability by about 40%.
The authors proposed a tetracycline-like chelate formation as the underlying
mechanism for this drug-drug interaction, where cefdinir chelates with iron via
its 7-hydroxyimino radical and its carbonyl oxygen, forming a poorly soluble
complex. A schematic of the complexation and the bioavailability data with
and without iron treatment are shown in Fig. l.
The coadministration of metal ions present in antacids, iron formulations
and nutrients such as dairy products with drug formulations might therefore
affect the therapeutic efficacy of the drugs. An 80%-90% reduction in plasma
levels of chlortetracycline was observed when the drug was coadministered
with aluminum-containing antacid gels, resulting in concentrations below the
therapeutic level (WAISBREN and HUECKEL 1950). Dairy products have been
shown to reduce dimeclocycline bioavailability by up to 80% (SCHREINER and
ALTEMEIER 1962).
However, in addition to solubility reduction as a consequence of chelate
formation, tetracycline dissolution might become impaired in the presence of
other compounds, depending on the formulation. Barr et al. have reported
that sodium bicarbonate, which lacks polyvalent cations, reduces tetracycline
absorption by 50% when the antibiotic was given as a capsule, but had no
effect when tetracycline was administered as a solution (BARR et al. 1971). This
was attributed to a reduced dissolution.
Another group of drugs reported to be affected by complexation with
metal ions are the anticoagulants. In contrast to the examples discussed
above, bishydroxycoumarin (BHC), for example, exhibits an increase in
absorption after oral coadministration together with milk of magnesia
(AMBRE and FISCHER 1972). Surprisingly, no effect was noted after aluminum
coadministration. A Mg-BHC chelate has been described previously, and
administration of the chelate to two subjects resulted in significantly increased
plasma levels, suggesting the preferential absorption of the chelate compared
to the parent drug. However, the mechanism by which the chelate increases
BHC plasma levels is unknown and whether the chelate is absorbed intactly or
releases BHC before absorption remains to be clarified. This interaction
potentially results in severe undesired pharmacological effects, considering
the low therapeutic index of this class of compounds.
As a final example, an up to tenfold reduction in absorption of quinolones
after antacid administration has been reported and has been primarily attrib-
uted to formation of chelate complexes (LODE 1988).
20 E. LIPKA et al.

II. Binding to Resins

Significant reduction in warfarin plasma levels of up to 30% after oral
coadministration with cholestyramine has been noted in a clinical trial with
six subjects (ROBINSON et al. 1970). Since patients with hyperlipidemia are
quite frequently treated with a cholesterol-lowering combination therapy
that includes cholestyramine and might also receive anticoagulants, such inter-
actions are again clinically important due to the narrow therapeutic interval of
anticoagulant agents. In vitro binding studies indicate a significant binding of
warfarin to cholestyramine at a pH above its pKa of 5.5. This suggests that the
bioavailability of warfarin is reduced as a result of a smaller fraction being
available for absorption. Furthermore, due to reduced lipid absorption, vita-
min K uptake is impaired, which makes the overall pharmacological response
highly unpredictable.
Highly stable binding of thyroxine, another drug with a low therapeutic
index, to cholestyramine has also been suggested as the underlying mechanism
for the reduced oral bioavailability when the two drugs are combined in the
treatment. A time interval of at least 4h between administration of the two
drugs is recommended (NORTHCUTI et al. 1969).

III. Complexation with Bile Salts

The absorption of nadolol has been shown to be strongly inhibited by tri-
hydroxy, but not by dihydroxy, bile salts (Fig. 2) (YAMAGUCHI et al. 1986a,b).

•
~

20
!
...t
~
III
QI
~
c:
...t
....
....III 10 4)

e
III
4)
4)

~ 4)
4)
QI

-=Ilo
~
0
tl
0 10 20 40
Concentration of bile salts (nut)

Fig. 2. Effect of bile salts on nadolol absorption 4 h after injection into the rat jejunum
loop. (Adapted from YAMAGUCHI et al. 1986b). closed circles, sodium cholate; closed
squares, sodium glycocholate; closed triangles sodium taurocholate; open circles, so-
dium chenodeoxycholate; open squares, sodium deoxycholate; open triangles, sodium
lithocholate. Dose of nadolol was 0.01 mmol/rat, and that of bile salts was 0.005-
O.lmmollrat. Each point represents the mean value for at least four rats. a) P < 0.01
compared to the control (without bile salts)
Drug Interactions in the Gastrointestinal Tract 21

Furthermore, no absorption inhibition was noted for other ,B-blocking agents,

implying that specific structural features are responsible for this interaction.
Subsequent reports demonstrated that nadolol forms a tenfold more stable
micellar complex with trihydroxy bile salts than other ,B-blockers (YAMAGUCHI
et al. 1986b). This complex was poorly soluble in water, resulting in a signifi-
cant decrease in nadolol uptake from the intestine. After performing magnetic
resonance imaging (MRI) spectroscopic studies with a series of nadolol de-
rivatives, the authors concluded that the cis-2,3-diol moiety of the molecule is
crucial for stabilizing the micellar complex (YAMAGUCHI et al. 1986c).

IV. Nonspecific Adsorption

Adsorption of therapeutic drugs to so-called inert excipients in formulations
as well as adsorption to charcoal administered for diarrhea therapy might
significantly interfere with the chronic treatment of patients. Rifampicin, for
example, is prescribed in a capsule formulation that contains bentonite, which
adsorbs the drug to a considerable extent (BOMAN et al. 1975). Consequently,
rifampicin bioavailability is reduced by 50% compared to administration of
the bentonite-free tablet.
An outbreak of anticonvulsant intoxication was observed in epileptic
patients in Australia during 1968-1969 (TYRER et al. 1970). The clinical picture
of all 51 patients was consistent with the diagnosis of anticonvulsant intoxica-
tion and it could be shown that all patients took lOO-mg sodium phenytoin
(diphenylhydantoin) capsules produced by one particular manufacturer. The
excipients in this capsule formulation had been changed from calcium sulfate
dihydrate to lactulose. This results of subsequent investigations showed that
the patient's blood phenytoin concentration fell rapidly to one-fourth of
its former value when the phenytoin capsules containing calcium sulfate as
excipients were administered instead of phenytoin capsules with lactulose.
The blood concentration rose again when the capsules containing lactulose as
excipient were administered. This phenomenon was of relevant therapeutic
importance, but it is not completely understood whether it is the result of
unspecific adsorption, complexation with metal ions, or other physicochemical
factors.
The inhibitory effect of charcoal on absorption was demonstrated in two
studies where the oral bioavailability of digoxin, phenytoin, phenobarbitone,
carbamazepine and phenylbutazone was reduced by more than 95%
(NEUVONEN et al. 1978; NEUVONEN and ELONEN 1980). In addition, a reduced
plasma half-life was reported for these compounds, which is most likely due to
an interruption of the enterohepatic circulation.
22 E. LIPKA et al.

c. Interactions with Drugs That Influence GI Motility

and pH
I. Alteration of Gastrointestinal Motility
Drugs that influence gastrointestinal motility and therefore transit time can
have a significant effect on drug uptake. Since very few drugs are absorbed
from the stomach, gastric emptying rate as well as small intestinal transit time
may become the rate-limiting steps to absorption. The effect of drugs that alter
intestinal pH and motility upon other drugs depends on the physicochemical
as well as the biological properties of these co administered compounds. A
decrease in motility might result in an increased extent of absorption as a
result of a prolonged dissolution time and increased contact time with the
intestinal surface. Alternatively, a decrease in absorption rate could result as
a consequence of delayed gastric emptying and low pH instability. The follow-
ing examples will illustrate some of the mechanisms described.
Propantheline, an anticholinergic drug, decreases motility and thus the
rate of gastric emptying, whereas the dopamine antagonist metoclopramide
increases gastric emptying rate. NIMMO et al. (1973) demonstrated the effect
of both compounds on acetaminophen absorption in healthy volunteers.
Propantheline decreased whereas metoclopramide increased the rate of acet-
aminophen absorption, correlating well with the simultaneously observed
changes (decrease and increase, respectively) in gastric emptying. However,
the overall extent of absorption as reflected by the urinary recovery did not
change, suggesting that small and large intestinal transit did not change.
A marked influence of propantheline and metoclopramide on the
absorption kinetics of ethanol has also been reported (GIBBONS and LANT
1975). Rapid gastric emptying induced by intravenous administration of
metoclopramide led to a close to 100% increase in peak plasma levels of
ethanol. The increased rate of absorption in combination with the saturable
first-pass metabolism of ethanol might result in severe undesirable side effects
if these compounds are administered simultaneously, e.g., if a formulation
contains significant amounts of ethanol. Intravenous administration of
propantheline significantly decreased the absorption rate of ethanol and
resulted in a 50% reduction of peak plasma levels.
The effects of these motility-modulating compounds on the absorption of
digoxin are based on a different mechanism. Here increased plasma levels
of digoxin after propantheline and decreased levels after metoclopramide
administration have been demonstrated (MANNINEN et al. 1973). Digoxin
uptake is rate limited by the drugs' solubility as well as the disintegration
and dissolution of the dosage form. Prolonged gastrointestinal residence time
induced by propantheline appears to result in a greater fraction of the
drug dissolved and consequently in an increased extent of absorption. The
opposite effect can be observed after metoclopramide treatment, where
impaired extent of drug absorption occurs as a result of the increased rate of
gastrointestinal transit.
Drug Interactions in the Gastrointestinal Tract 23

These observations can be theoretically described employing a simple

mass balance model that represents the intestine as a tube to predict extent
of absorption as a function of dissolution and permeability parameters (Ho
et al. 1983; NI et al. 1980). In order to develop a more quantitative and
predictive model for drug absorption rates, it is necessary to include in the
model flow, dissolution, absorption and reaction processes occurring in the
intestine. In general this is quite complex. However, a simple model that
considers a segment of intestine over which the permeability may be consid-
ered constant, a plug flow fluid with the suspended particles moving with the
fluid, no significant particle-particle interactions (i.e., aggregation) and disso-
lution in the small particle limit, leads to the following pair of differential
equations in dimensionless form (AMIDON et al. 1995):
dr* / dz * =-( Dn/3)(1- C*)/ r* (1)
and
dC* / dz * = DoDnr* (1- C*) - 2AnC* (2)
where
z* = z/L = (vz!L)t = t*
= t/(Llv z ) = t/(AL/Q) = t/(V/Q)
t*
where L = tube length, Vz = axial fluid velocity in the tube, A = tube surface
area = 2nRL, R = tube radius, Q = axial fluid flow rate = A Vz • The three
important dimensionless groups are:
MolVo
Do = dose number = --"-'---=-
Cs

D n -- d·1SS0lutlOn DC
· num b er -- - _s. -4Jrr~ I
- - . tres -- tres . 3DC s pro2
ro 4 3
-Jrr
3 0P

An = absorption number = P;f . tres = t;~s . tres

tres = nR2 L/ Q = mean residence time

t Diss = -r~-P
- =.tIme reqUIre
. d f or a partIe
. 1e to d·1SS01ve
3DC s

1
t~bs = k abs = ( S /V) Peff = 2 xRPeff = the effectIve
. absorptlOn
. rate constant

It is clear from the definition of the dissolution number that if the dissolution
rate and absorption rate in the intestine are constant (constant permeability,
24 E. LIPKA et al.

1.0

:f 0. 5

Fig. 3. Fraction dose absorbed versus dose and dissolution number

particle size and solubility, e.g., a nonpolar non-ionizable drug), gastrointesti-

nal residence time will be the primary source of variation. Figure 3 is a plot of
the fraction dose absorbed versus the dose and dissolution number for drugs
with low solubility and high permeability estimated by Eqs. 1 and 2. The most
important compounds affected by transit time are those that fall on the steep
area of the surface, i.e., dose number", 1-10 and a dissolution number'" 0.1-
10. For these drugs, small changes in transit time can cause large changes in the
fraction absorbed. In general, the average small intestinal transit time (SITT)
is 3.2 ± 1.3h, whereas the large intestinal transit time (UTI) is 32 ± 18h and
represents a variability inherent in the population (DAVIS et al. 1986). Refer-
ring back to digoxin as an example, which has an aqueous solubility of
0.024mglml and a dose of 0.5mg, one can expect about 50% absorption,
assuming an average particle diameter of 50.um. Changing the dissolution
number by small increments, either positively or negatively, will result in large
changes in the fraction dose absorbed. JOHNSON et al. (1978) studied the effect
of transit time and particle size on the absorption of digoxin. Their study
showed that the micronized product (dave< lO.um) was not affected by changes
in the intestinal transit time whereas the unmicronized product (dave'" IOO.um)
was significantly affected. For the 100-.um particles, the cumulative 4-day
urinary excretion of digoxin was increased by 14 % when co administered with
propantheline (reduces GI motility and transit time) and reduced by 27%
Drug Interactions in the Gastrointestinal Tract 25

0.25

0.20 F '" 0.20

E 0.15
c ~
.;(0 8....
elli
=a- .~ 0.10
LI-'a

F '" 0.06
0.05

0.00 +---....---....---....---....----.
o 10 20 30 40 50
Percent Deviation from Mean SITT

Fig. 4. Predicted fraction dose absorbed for lOO-.um particles of digoxin at different
small intestinal transit times (SlIT)

when co administered with metoclopramide (increases GI motility and transit

time). The predicted fractions of the dose absorbed for different transit times
are shown in Fig. 4. The model predicts 100% absorption for the micronized
digoxin.
The existence of an "absorption window" in the upper part of the
small intestine represents the third hypothesis discussed as the underlying
mechanism of drug interactions with transit time altering compounds.
RESETARITS and BATES (1979) reported a possible absorption-enhancing effect
of propantheline bromide on orally administered chlorothiazide in dogs,
which they attributed to a longer residence time at the absorption site.
Based on the relatively low solubility of chlorothiazide at pH values
comparable to small intestinal fluids (pH 6.2) of 0.65mg/ml, however,
the authors also considered an increase in dissolution time as a possible
explanation. Furthermore, chlorothiazide exhibits a dose-dependent decrease
in absorption at higher doses, which could be in part reversed by simultaneous
administration of propantheline bromide, again suggesting that dissolution
might be the rate-limiting step to chlorothiazide absorption. Increased urinary
excretion of cholorothiazide after propantheline administration and decreased
excretion after metoclopramide administration has also been reported in
humans, with saturable Or site-specific absorption as well as solubility
limitations being considered as possible mechanisms (OSMAN and WELLING
1983). Simulation studies with chlorothiazide that included drug parameters
such as pKa' solubility and intestinal intrinsic permeability as well as varying
physiological parameters (pH profile, volume of intestinal contents and
intestinal flow rates) also indicated that the dose dependency of chlorothiazide
26 E. LIPKA et al.

may be attributed to solubility-ionization effects rather than a saturable

absorption mechanism (DRESSMAN et al. 1984).
One of the few examples where prolonged gastric residence time as a
result of anticholinergic drug administration leads to reduced oral availability
is L-dopa. Both trihexylphenidyl and atropine significantly lower plasma levels
of L-dopa after oral administration (ALGERI et al. 1976). Metabolism by the
gastric mucosa was reported for L-dopa; therefore a delay in gastric emptying
might result in decreased oral bioavailability of the parent drug in this case
(RIVERA-CALIMLIM et al. 1970a-c). Alternatively very slow gastric emptying
could lead to higher small intestinal metabolism if the metabolism is nonlinear
and/or higher in the upper small intestine.
It is widely known that ingestion of ethanol influences gastrointestinal
motility. Significantly delayed gastric emptying has been shown, e.g., as a
result of administration of highly concentrated alcoholic beverages
(BARBoRIAK and MEADE 1970). However, other possible mechanisms have also
been discussed for the ethanol-influencing bioavailability of certain drugs.
Thus, ethanol may modify/affect solubility of the drug substance itself.
Furthermore, permeability of absorptive membranes might be increased
(WHITEHOUSE et al. 1975). In a two-period crossover study, HAYES et al. (1977)
demonstrated that bioavailability of diazepam was significantly increased
when the drug was co administered with 30ml of 50% ethanol. In this study
maximum plasma concentration as well as extent of bioavailability were
significantly higher. Comparable effects were reported for sulfapyridine,
chloral hydrate and chlordiazepoxide (RIETBROCK et al. 1983).
II. pH Alteration in the Gastrointestinal Tract
Aluminum hydroxide, widely used in a variety of antacid formulations, repre-
sents an intermediate compound in terms of its effects on the gastrointestinal
system. As an antacid, aluminum hydroxide increases gastric pH; in addition it
also delays gastric emptying. Thus, it could affect drug absorption via two
possible mechanisms. HURWITZ (1971) investigated the effects of aluminum
hydroxide and magnesium hydroxide on the oral absorption of sulfadiazine,
a weak acid, and quinine, a weak base, in rats. After pretreatment
with aluminum hydroxide, administration of sodium sulfadiazine resulted in
significantly lower plasma levels, which was attributed to a delay in gastric
emptying, whereas magnesium hydroxide did not alter the plasma profile. In
contrast, an increase in plasma levels was reported when sulfadiazine was
administered as the free acid after pretreatment with magnesium hydroxide. It
was proposed that magnesium hydroxide raises gastric pH sufficiently to
increase the solubility of the free acid. Both magnesium hydroxide and
aluminum hydroxide decreased quinine absorption, the former predominantly
by raising gastric pH and precipitating the quinine, the latter additionally by
delaying gastric emptying.
Decreased oral bioavailability of quinolone antibiotics after simultaneous
administration of antacids has been widely reported. Aluminum-magnesium
Drug Interactions in the Gastrointestinal Tract 27

Table 2. Mean ± SD values of enoxacin pharmacokinetic parameters alone and after

coadministration with ranitidine, pentagastrin and ranitidine + pentagastrin. (Adapted
from LEBSACK et al. 1992)

Treatment Cmax (Jig/ml) t max (h) AVC (ugh/ml) t1l2 (h) Cl, (ml/min)

Enoxacin alone 2.74 ± 0.68 1.3 ± 0.5 16.1 ± 3.6 5.2 ± 1.0 182 ± 73
Enoxacin + 1.70 ± 0.66* 2.0 ± 1.0 11.9 ± 5.2* 5.7 ± 1.0* 211 ± 46
ranitidine
Enoxacin + 2.49 ± 0.55 2.1 ± 0.9 16.2 ± 4.9 5.2 ± 0.8 198 ± 81
pentagastrin
Enoxacin + 2.93 ± 0.66 1.5 ± 0.5 17.7±5.6 5.5 ± 0.9* 212 ± 86
ranitidine +
pentagastrin

Cmax> peak plasma concentration; tmaX' time to reach peak concentration; AVC, area
under the plasma concentration-time curve; (liZ, half-life; Cl" renal clearance.
* Significantly different from enoxacin alone (P < 0.05).

hydroxide, for example, reduced the mean oral bioavailability of norfloxacin

and ciprofloxacin by 91% and 85%, respectively, which was previously
attributed to chelate complexation (NIX et al. 1989, 1990). However, earlier
studies demonstrated a 40% decrease in enoxacin bioavailability also after
ranitidine pretreatment, therefore indicating that increased gastric pH could
also be a factor when antacids are co administered with quinolones (GRASELA
et al. 1989). Recently, LEBSACK et al. (1992) reported a 26% decrease in the
oral bioavailability of enoxacin after intravenous administration of ranitidine
(Table 2). This effect could be reversed by the simultaneous administration of
pentagastrin, suggesting that a decrease in gastric acidity rather than a direct
interaction of enoxacin with ranitidine is responsible for this interaction. This
was further supported by the pH-dependent aqueous solubility of enoxacin.
At pH values below 4.5, enoxacin is highly soluble (up to 16mg/ml at pH
2.0), whereas its solubility drops below 1 mg/ml at pH values above 5. The
pH-dependent solubility might therefore be the rate-limiting step to enoxacin
absorption. However, to what extent antacids as well as H 2 -antagonists alter
the pH in the jejunum and ileum, the major absorption sites, and possibly
influence reprecipitation of the drug remains to be evaluated.
Bioavailability of drugs may also be influenced by coadministration of
H 2-antagonists when dissolution of the active ingredient from the formulation
is pH-dependent. This was shown for a controlled/modified-release verapamil
formulation where in vitro dissolution experiments showed markedly reduced
dissolution in media buffered at pH 8.0, 6.8 or 4.5 compared to pH 1.2 (BLUME
1990; BLUME et al. 1988). A comparative bioavailability investigation of
this product showed that intravenous coadministration of famotidine resulted
in a reduction of bioavailability of verapamil during the first 2h after
administration, indicating an effect on the dosage-form-dependent release
rate due to an increase of gastric pH.
28 E. LIPKA et al.

:I
i
.!!
:.0
>.~
.c'O
"0
~ a
t:

~~0
0.:;
cb\(

1
o~

.;
<

o~~--~--~~--~--~--~~--~
o 20 40 60 80
Absorbability of model drugs
(% absorbed in 1 h)

Fig. 5. Plot of the effect of bile salts on drug absorption against the absorbability of
the model drugs. (Adapted from KIMURA et al. 1985). Intestinal absorption of drugs
was examined in the presence of 20mM bile salts or sodium lauryl sulfate. Each
point represents the mean of at least three experiments. Drugs: 1, sulfaguanidine;
2, sulfanilamide; 3, sulfadimthoxine; 4, sulfanilacetamide; 5, sulfisoxazole; 6,
sulfapyridine; 7, sulfamethoxypyridazine; 8, sulfaphenazole; 9, sulfisomidine; 10,
sulfamethizole; 11, phenol red; 12, imipramine; 13, quinine; 14, 2-allyloxy-4-chloro-N-
(2-diethylaminoethyl)-benzamide; 15, NI,NI-anhydro-bis-(fJ-hydroxy-ethyl)biguanide;
16, metoclopramide; 17, procainamide. Filled circles, with sodium taurocholate; open
circles, with sodium cholate; filled triangles, with sodium taurodeoxycholate; open
triangles, with sodium lauryl sulfate

D. Interactions Between Drugs

That Share the Same Absorption Mechanism

I. Passive Absorption
Various compounds with proposed enhancing effects on passive transcellular
as well as paracellular absorption have been investigated over recent years,
including oleic acid, sodium lauryl sulfate and bile salts (MURANASHI 1990).
KIMURA et al. (1985) summarized in a detailed report the effects of di- and
trihydroxy bile salts on the in situ absorption of a broad range of compounds
as depicted in Fig. 5. They proposed different mechanisms with regard to both
the observed enhancing as well as inhibitory effects of bile salts on drug
absorption. Sodium taurodeoxycholic acid (STDC) interacts most likely di-
rectly with the intestinal membrane by forming mixed aggregates with the
phospholipids in the lipid bilayer, therefore altering the membrane structure.
Drug Interactions in the Gastrointestinal Tract 29

Compounds that are poorly absorbed by the unperturbed membrane usually

exhibit an increased rate of absorption in the presence of bile salts. Further-
more, both STDC and STC (sodium taurocholic acid) appear to induce cal-
cium depletion of the tight junctions, resulting in an opening of the cell
junctions, therefore facilitating paracellular diffusion of small polar drugs.
Well-absorbed compounds, in contrast, may exhibit a decrease of their absorp-
tion rate when administered with bile salts. This observation might be in part
due to a reduction in thermodynamic activity as a consequence of micelle
formation as well as to depletion of some mucosal components, lipids and/or
proteins to which the drug may have a high affinity.
Interactions of drugs that are absorbed via the paracellular pathway have
been rarely described. As mentioned above, an opening of the tight junctions
due to calcium depletion induced by bile salts might lead to an enhanced
uptake of small hydrophilic compounds. Another mechanism where opening
of the intercellular space facilitates drug uptake is what is commonly known
as solvent drag effect. Electrolytes and glucose induce convective transport
of other molecules by increasing the osmolarity in the aqueous channels
between the cells and therefore increasing net water uptake. The intestinal
permeability of acetaminophen, a small water-soluble drug, in rats increased
significantly as a function of glucose concentration (Fig. 6) (FLEISHER 1995).
The solvent drag approach was also utilized as an attempt for increasing the
oral bioavailability of small peptides and peptide analogues. Permeability
increase in the presence of glucose was demonstrated for the dipeptide D-
kytorphin and the tripeptide analogue cephadrine (FLEISHER 1995). However,
no effect of glucose was observed for the uptake of a growth hormone releas-
ing hexapeptide and a cyclic octapeptide somatostatin analogue, which
probably reflects the size cut-off of the paracellular route. Studies in humans
of the effect of water transport on drug absorption are relatively few, but
recent permeability studies suggest that there is only a limited effect. In
fact, there appears to be evidence that D-glucose as well as L-Ieucine fail to
stimulate water absorption in the human jejunum, leading to a reevaluation of
the "solvent drag" concept in humans (LENNERNAS 1995).

II. Carrier-Mediated Absorption

The amino acid carrier as well as the small peptide carrier, which transports
di- and tripeptides, are two well-characterized active transport systems in the
small intestine (MAILLARD et al. 1995; WILLIAMSON and OXENDER 1995). While
the former exhibits a high substrate specificity, the latter is known to facilitate
the transport of a diverse group of peptides and peptide analogues, which
might result in possible interactions during absorption. Compounds that are
supposed to be absorbed via the small peptide carrier include amino f3-lactam
antibiotics and angiotensin-converting enzyme (ACE) inhibitors (TSUJI 1995;
YEE and AMIDON 1995).
Representative drugs suggested to be absorbed by the amino acid
pathway are L-dopa and a-methyldopa. Studies using L-Ieucine as an inhibitor
30 E. LIPKA et al.

Fig. 6. Acetaminophen (APAP) permeability correlates with jejunal water absorption

induced by glucose in rat perfusion, suggesting solvent drag of drug through
paracellular pathways. (Adapted from FLEISHER 1995)

of L-dopa uptake in single-pass perfusion studies in rats demonstrated a ten-

fold decrease in intestinal wall permeability, indicating transport of L-dopa by
the amino acid carrier (Fig. 7) (SINKO et al. 1987). LENNERNAS et al. (1993)
provided evidence in humans for a competitive uptake via this carrier by
demonstrating a decrease in L-dopa intestinal absorption from 40% to 21 % in
the presence of L-Ieucine.
Cefatrizine absorption was significantly inhibited by the dipeptide
Phe-Phe, whereas L-Phe had no effect on its absorption, suggesting that
cefatrizine uptake occurs via the small peptide transporter (SINKO et al. 1987).
Drug Interactions in the Gastrointestinal Tract 31

0>-
f-
..... «
-1Q.
..... 0
CDO
« I 4 T
W-1
~
a:l.L.
WO
1
Q.
T
2 1
_T
1 T
.!. T

1 I
.J..

o
o 10 25 50 75 100
CONCENTRA nON LEUCINE (mM)

Fig. 7. Plot of the effect of the inhibitor L-Ieucine on the wall permeability of L-dopa.
The concentration of L-dopa remained constant at 0.1 mM whereas the concentration
of L-Ieucine varied from 0 to 100mM. (Adapted from SINKO et al. 1987)

Mutual absorption inhibition among amino f3-lactam antibiotics was also

hypothesized as a result of competition for the small peptide transporter
(ISEKI et al. 1984), but conflicting results were obtained from rat perfusion
studies. The absorption of amoxicillin was significantly inhibited by cyclacillin,
cephadrine and cephalexin, but no effect of these compounds on ampicillin
absorption was observed. Furthermore, cephalexin absorption was reduced
by L-carnosine, but not by glycylglycine, and cephadrine uptake was not
affected by any of the dipeptides investigated. While there appears to be
considerable evidence that f3-lactam antibiotics are transported by the small
peptide carrier, there is still no general agreement about the absolute number
of transporters and their substrate selectivity (WALTER et al., accepted). Thera-
peutically significant interactions of drugs that are primarily absorbed
by this carrier system have not been demonstrated to date, but further
elucidation of the substrate specificity and selectivity of this transport
pathway(s) is required.
A clear inhibition of the absorption of the ACE inhibitor captopril by
various compounds including dipeptides and cephadrine was evident from in
situ perfusion studies in rats (Table 3) (Hu and AMIDON 1988). However, a
clinical study in humans failed to show a significant effect of cephradine
coadministration on captopril bioavailability (Yee and Amidon, unpublished
32 E. LIPKA et al.

Table 3. Effect of various compounds on captopril permeability. (After Hu and

AMIDON 1988)

Compound coperfused with Pw* (SEM) P value P value P value

captopril (1 mM) (to A)' (to B) (to C)

A = none 2.63 (0.1) _b

B = 5mM Pro-Phe 2.22 (0.34)

C = NAP and NAC (50mM) 2.19 (0.43)
D = 85 mM Gly-Gly 1.10 (0.27) 0.005 0.1 0.1
E = 5 mM 2,4-DNP 0.64 (0.14) 0.001 0.025 0.025
F = lOmM cephadrine 0.45 (0.15) 0.001 0.025 0.05
G = no sodium 0.98 (0.18) 0.005 0.05 0.01
H = 15 mM Gly-Pro 0.65 (0.22) 0.001 0.025 0.05
I = peptide mixture c 0.58 (0.09) 0.001 0.025 0.025

PW*, dimensionless intestinal wall permeability.

'Two-sample (-test performed comparing condition A with other conditions (B, C,
etc.): P < 0.05, significant; P < 0.01, very significant; P > 0.05, not significant.
b Not statistically different.
c A mixture of 2mM each of Gly-Pro, Pro-Phe, Asp-Phe and Pro-Tyro

results). Since there is a significant passive component in the absorption of

captopril, conclusions with regard to clinically important interactions between
drugs transported by the dipeptide carrier must be tentative.
P-glycoprotein (Gp 170) is a glycoprotein coded by the multidrug
resistance gene (MDR), which was first described in cancer tissue, where
it served as an efflux transporter for exogenous compounds including
chemotherapeutics, leading to drug resistance and therapy failures. Recently
the existence of P-glycoprotein was also demonstrated in normal tissue such as
the blood-brain barrier, the kidneys, the liver and the jejunum (CORNWELL
1991). Furthermore, it was suggested that the transporter might protect
against exogenous compounds and contribute to the rarity of small intestinal
cancer in humans, a hypothesis that needs further investigation (HSING et al.
1992).
A variety of therapeutic drugs are substrates for this secretory carrier,
with the chemotherapeutics vinblastine and vincristine and the calcium chan-
nel blocker verapamil being the most intensely studied drugs (INCE et al. 1988).
Early work suggested a competitive interaction at the transport protein, lead-
ing to an increased antitumor efficacy of vincristine in the presence of
verapamil. More recently, HUNTER et al. (1993) demonstrated a marked
increase of vinblastine absorption after simultaneous administration of
verapamil in Caco-2 cells. The significant efflux of vinblastine from the
basolateral to the apical side of the cell model could be inhibited by verapamil,
whereas the almost non detectable flux of vinblastine from the apical to the
basolateral side was significantly increased after addition of the calcium
channel blocker.
It has been hypothesized that drugs other than chemotherapeutics might
be transported by P-glycoprotein, consequently offering an alternative
Drug Interactions in the Gastrointestinal Tract 33

Condition

Control

Vinblastine 0.5 mM
on APside ***
Vinblastine 0.5 mM
on BL side ***
Celiprolol 5 mM
on APside
Celiprolol 5 mM
on BL side
a 25 50 75 100 125
Basal-to-apical [3H)-vinblastine transport (%)

Fig. 8. Effect of unlabeled celiprolol and vinblastine on basal-to-apical transport of

[3H]-vinblastine in Caco-2 cell monolayers .The basal-to-apical transport of lOnM (3H]-
vinblastine was determined after 30min in the presence of excess unlabeled celiprolol
and vinblastine on the basal (BL) or apical (AP) side. Data are presented as a percent-
age of control transport. Each bar represents mean ± SD; n = 4. Significant differences:
***p < 0.001. (Adapted from KARLSON et al. 1993)

explanation for the frequently observed nonlinear oral absorption kinetics of

drugs. Examples of such drugs include celiprolol, pafenolol and talinolol, three
hydrophilic ,B-blockers that exhibit a dose-dependent increase in their oral
pharmacokinetics. Celiprolol was shown to be secreted into the intestinal
lumen in a time-, temperature- and pH-dependent manner in rats (Kuo et al.
1994). Studies in Caco-2 cells indicated that celiprolol secretion could be
inhibited by vinblastine and verapamil, further supporting the involvement of
P-glycoprotein in the secretory mechanism (Fig. 8) (KARLSON et al. 1993). For
pafenolol, significant intestinal secretion in rats has been reported as a possible
explanation for the nonlinear oral bioavailability (LENNERNAS and REGARDH
1993). A very recent study with talinolol in humans as well as inhibition
studies in Caco-2 cells also suggests that the variable and dose-dependent
oral absorption of this drug may be due to a potential interaction with P-
glycoprotein (WETIERICH et al. 1995). However, since the Caco-2 cell line is a
transformed colon cell line, the extent to which this transport protein might be
involved in clinically significant drug interactions in vivo remains to be further
investigated.

E. Interactions as a Function of Intestinal Metabolism

Numerous in vitro and in vivo studies have demonstrated that many drugs and
xenobiotics are metabolized by the gastrointestinal tract, either in the gut
lumen or in the intestinal wall, resulting in high inter- and also intraindividual
variations in oral bioavailability (ILETI et al. 1990). Metabolic enzymes in
the lumen originate from exocrine glands and are predominantly active in the
upper parts of the intestine (RENWICK 1982; KRISHNAMOORTHY and MITRA
34 E. LIPKA et al.

1995). Enzymatic reactions as a function of bacterial organisms include degra-

dation, reduction and hydrolysis and take place in the terminal portions of the
gut. A variety of enzymatic reactions also occur in the intestinal mucosal cell.
These include phase I biotransformations such as oxidation, reduction and
hydrolysis and several phase II conjugation reactions, with glucuronidation
and sulfation representing the most prominent ones. While the concentrations
of phase I enzymes were generally assumed to be relatively low, conjugative
phase II enzyme activity is comparable to that in the liver (TAM 1993).
Some examples with regard to drug interactions can be found in the early
literature, such as the undesired increase in ethinylestradiol plasma levels after
concomitant administration of either acetaminophen or ascorbic acid, two
compounds that compete with ethinylestradiol for the limited sulfate pool
(BACK and ROGERS 1987). However, the extent of interactions in the gut wall
might have been overestimated in these studies, since liver contribution was
not taken into account. These considerations might also be an explanation for
interactions reported for morphine and fenoterol, as well as morphine and
orciprenaline (KOSTER et al. 1985a,b).
The contribution of intestinal phase I metabolism was regarded as un-
likely to be important until recently, since the ratio of hepatic to intestinal
cytochrome P450 concentration has been reported to be approximately 20
(TAM 1993). However, recent reports focus on phase I biotransformations,
predominantly via cytochrome P450, in the intestinal wall and provide a more
quantitative and mechanistic understanding of these metabolic reactions.
WATKINS et al. (1987) were the first to report that a major cytochrome P450,
CYP3A, which appears to account for approximately 70% of the total P450
present in human enterocytes, might be present in a higher concentration in
the intestine than in the liver. This high level of a cytochrome P450 in the
intestine may be even more important considering that more than 50% of
drugs used in humans might be substrates for this enzyme (WACHER et al.
1995). Consequently, metabolic interactions could be very likely. In subse-
quent studies the same authors reported that cyclosporin, which was assumed
to be metabolized exclusively in the liver, undergoes at least 50% metabolism
in the intestinal wall as measured from portal and femoral blood in anhepatic
liver transplant patients (Table 4) (KOLARS et al. 1991). Considering that not
all metabolites of cyclosporin were measured in this study, the contribution of
intestinal metabolism might be even higher. Furthermore, intestinal CYP3A
has been shown to be inducible by rifampicin, leading to an abolished oral
bioavailability of cyclosporin in another transplant patient studied (KOLARS
et al. 1992). GOMEZ et al. (1995) confirmed these findings in a study where
ketoconazole, a CYP3A inhibitor, induced an increase of cyclosporine
bioavailability from 22.4% to 56.4%. Grapefruit juice and its possibly active
component, naringin, have been used for investigating interactions of drugs
metabolized by CYP3A. Felodipine area under the plasma concentration-time
curve (AUe) and peak plasma concentrations increased about 200% and
170%, respectively, when co administered with grapefruit juice, whereas
Drug Interactions in the Gastrointestinal Tract 35

Table 4. Cyclosporin and metabolite (M1 and M21)

concentrations in portal vein (PV) and femoral artery (FA)
after duodenal instillation during anhepatic phase.
(Adapted from KOLARS et al. 1991)

Time after Cyclosporin M1 + M21

instillation (min) (ng/ml) (ng/ml)

PV FA PV FA

Patient N
0 0 0 0 0
30 56 37 12 16
60 124 92 42 32
Pateint B
0 0 0 0 0
40 14 29 0 0
120 22 18 23 10

a In patient A the cyclosporin solution was mixed with 12 ml

bile before instillation to facilitate absorption.

naringin produced much less of an effect in this study (Fig. 9) (BAILEY et al.
1993). AUC ratios between dehydrofelodipine (the main metabolite of
felodipine) and felodipine were reduced, suggesting inhibition of presystemic
metabolism. The relative contribution of intestinal metabolism in contrast to
hepatic metabolism remains to be determined for felodipine.
For cyclosporine (DUCHARME et al. 1995) and midazolam (KUPFERSCHMIDT
et al. 1995) on the other hand, it was demonstrated that grapefruit juice had no
effect on pharmacokinetic parameters such as AUC, systemic clearance and
elimination half-life when the drugs were administered intravenously. How-
ever, after oral administration of cyclosporine as well as midazolam together
with grapefruit juice an increase in bioavailability of 50% and 30%, respec-
tively, has been observed. The lack of effects on systemic clearance after
intravenous dosing suggests that oral bioavailability is improved by either
increased absorption or inhibition of gut-wall first-pass metabolism, with the
latter being more likely considering previous reports (KOLARS et al. 1991,
1992). However, one cannot completely rule out the effect on gastrointestinal
transit and increased solubilization due to bilary secretions as also being
partial explanations of these observations.
Clinically significant and severe interactions between azole antifungal
drugs and terfenadine, both substrates for CYP3A, have been reported
(HONIG et al. 1993; POHJOLA-SINTONEN et al. 1993). The decreased metabolism
of terfenadine to its active acid metabolite in the presence of either
itraconazole or ketoconazole results in increased plasma levels of
unmetabolized terfenadine. Terfenadine itself prolongs the QT interval and
increases the risk of torsade de pointes ventricular tachycardia. This metabolic
interaction might already occur on the intestinal level, but again the percent-
36 E. LIPKA et al.

15

-
0 0 WATER
:::;- • • GRAPEFRUIT JUICE
(5 )( )( NARINGIN SOL 'N
E
.s 10
0
Z
0
0
w
ii!: 5
9::
0
0
-'
W
IL

0
0.0 2.0 4.0 6.0 8.0

:::;- 20
~
E
.s
0 15
Z
0
0
w
~ 10
11.
0
0
-'
W
IL 5
0
a::
0
>
150
2.0 4.0 6.0 8.0

HOURS
Fig. 9. Plasma felodipine and dehydrofelodipine concentration-time profiles for the
three treatment groups receiving 5 mg felodipine with 200ml water, grapefruit juice or
naringin solution. Bars represent SEMs. Comparisons were made between results with
water and the two other treatments; *p < 0.025; **p < 0.01. (Adapted from BAILEY
et al. 1993)

age of intestinal metabolism in contrast to hepatic metabolism remains to be

determined. These combined findings suggest that this intestinal metabolic
pathway might be relevant for the low and highly variable oral bioavailability
of a large number of substrates of cytochrome CYP3A, including erythromy-
cin, lignocaine and estrogen.
There are few examples in the literature that illustrate pre systemic re-
moval of drugs after oral administration as a function of bacterial metabolism
in the distal intestine. The clinically most relevant interaction resulting
from decreased bacterial metabolism is the coadministration of estrogen and
Drug Interactions in the Gastrointestinal Tract 37

antibiotics. f3-Glucuronidase, f3-glycosidase and sulfatase in intestinal

microorganisms are responsible for hydrolyzing biliary-excreted estrogen
conjugates and allow enterohepatic circulation. Failure of oral contraceptives
has been associated with a significantly lower quantity of microorganisms in
the lower gastrointestinal tract after antibiotic treatment (BACK et al. 1978).
In contrast, digoxin has been demonstrated to be inactivated via enzymatic
reduction by the intestinal flora. This process is inhibited by antibiotic therapy,
resulting in an increased oral absorption (DOBKIN et al. 1983). It could be
demonstrated that a 5-day coadministration of erythromycin or tetracycline
markedly reduced the inactivation of digoxin by gut bacteria (LINDENBAUM
et al. 1981).

F. Altered Absorption as a Result

of Drug-Induced Mucosal Changes
Since administration of cytostatic drugs often results in pronounced changes in
the intestinal mucosa, the bioavailability of drugs may be affected by such
concomitant therapy. Bioavailability may be reduced in particular in the case
of lipophilic drugs, which are mainly passively absorbed. Such an effect is of
primary relevance for drugs with a small therapeutic range, e.g., digoxin. It
has been shown that coadministration of cytostatic drugs (combination of
cyclophosphamide, oncovin and procarbazine or cyclophosphamide, oncovin
and prednisone) diminished steady-state plasma concentration of digoxin
in patients with a daily oral dosage of O.3mg f3-acetyldigoxin (KUHLMANN
et al. 1981). In addition after single-dose administration of digoxin to
patients under cytostatic therapy the maximum plasma concentration was
significantly delayed. These results could be explained by changes in mem-
brane permeability and/or an effect on overall gastrointestinal transit.

References
Adjepon-Yamoah KK, Scott DB, Prescott LF (1973) Impaired absorption and
metabolism of oral lignocaine in patients undergoing laparoscopy. Br J Anaesth
45:143-147
Albert KS, Ayres JW, Disanto AR, Weidler DJ, Sakmar E, Halmark MR, StU RG,
DeSante KA, Wagner JG (1978a) Influence of kaolin-pectin suspension of digoxin
bioavailability. J Pharm Sci 67:1582-1585
Albert KS, DeSante KA, Welch RD, DiSanto AR (1978b) Pharmacokinetic evaluation
of a drug interaction between kaolin, pectin and clindamycin. J Pharm Sci
67:1579-1582
Algeri S, Cerletti C, Curcio M, Morselli PL, Bonollo L, Guniva G, Minazzi M, Minoli
G (1976) Effect of anticholinergic drugs on gastro-intestinal absorption of L-dopa
in rats and in man. Eur J PharmacoI35:293-299
Ambre JJ, Fischer U (1972) Effect of coadministration of aluminum and magnesium
hydroxides on absorption of anticoagulants in man. Clin Pharmacol Ther
14(2):231-237
Amidon GL, Lennernas H, Shah VP, Crison JR (1995) A theoretical basis for a
biopharmaceutics drug classification: the correlation of in-vitro drug product dis-
solution and in-vivo bioavailability. Pharmacol Res 12(3):413-420
38 E. LIPKA et al.

Back DJ, Rogers SM (1987) First-pass metabolism by gastrointestinal mucosa. Aliment

Pharmacol Ther 1:339-357
Back DJ, Breckenridge MC, Crawford FE, Challiner M, Orme ML (1978) The effect of
antibiotics on the enterohepatic circulation of ethinyl-estradiol and norethisterone
in rats. J Steroid Biochem 9:527-531
Bailey DG, Arnold JM, Munoz C, Spence JD (1993) Grapefruit juice-felodipine inter-
action: mechanism, predictability and effect of naringin. Clin Pharmacol Ther
53( 6):637 -642
Barboriak 11, Meade RC (1970) Effect of alcohol on gastric emptying in man. Am J
Clin Nutr 23:1151-1153
Barlow OW, Climenko DR (1941) Studies on the pharmacology of sulfapyridine and
sulfathiazole. JAMA 116:282-286
Barr WH, Adir J, Garrettson L (1971) Decrease of tetracycline absorption in man by
sodium bicarbonate. Clin Pharmacol Ther 12:779-784
Barrowman JA, D'Mello A, Herxheimer A (1973) A single dose of neomycin impairs
absorption of vitamin A (retinoil) in man. Eur J Clin Pharmacol 5:199-202
Beerman B, Groschinsky-Grind M (1978a) Enhancement of the gastrointestinal
absorption of hydrochlorothiazide by propantheline. Eur J Clin Pharmacol
13:385-387
Blume H (1990) Einfluss von Mahlzeiten auf die Bioverfuegbarkeit von
Retardarzneimitteln: sind "food-studies" erforderlich? In: Blume H (ed)
Bioverfuegbarkeit und Bioaequivalenz von Retardarzneimitteln; proceedings of
the 8th ZL-expert meeting, 20 March 1990, Govi, Frankfurt
Blume H, Siewert M, Eichelbaum M (1988) Vortrag anlaesslich des workshops "In-
vitro and in-vivo testing and correlation of controlled/modified release dosage
forms". 14-16 Dec 1988, Washington DC, USA
Boman G, Hanngren A, Malmborg A-S, Borga 0, Sjoqyist F (1971) Drug interaction:
decreased serum concentrations of rifampicin when given with P.A.S. Lancet
11:800
Boman G, Lundgren P, Stjernstrom G (1975) Mechanism of the inhibitory effect of
PAS granules on the absorption of rifampicin: adsorption of rifampicin by an
excipient, bentonite. Eur J Clin Pharmacol 8:293-299
Bradbrook ID, Morrison PJ, Rogers HJ (1978) The effect of bromelain on the absorp-
tion of orally administered tetracycline. Br J Clin Pharmacol 6:552-554
Brown DD, Juhl RP (1976) Decreased bioavailability of digoxin due to antacids and
kaolin-pectin. N Engl J Med 295:1034-1037
Caldwell JH, Bush CA, Greenberger NJ (1991) Interruption of the enterohepatic
circulation of digitoxin by cholestyramine. II. Effect on metabolic disposition of
tritium-labeled digitoxin and cardia systolic intervals in man. J Clin Invest
50:2638-2644
Cheng SH, White A (1962) Effect of orally administered neomycin on the absorption
of penicillin V. N Engl J Med 267:1296-1297
Chin T-F, Lach JL (1975) Drug diffusion and bioavailability: tertracycline-metallic
chelation. Am J Hosp Pharm 32:625-629
Cornwell MM (1991) Molecular biology of P-glycoprotein. In: Ozols R (ed) Molecular
and clinical advances in anticancer drug. Kluwer Academic, Boston, pp 37-
56
Crammer JL, Rosser RM, Crane G (1974) Blood levels and management of lithium
treatment. Br Med J 3:650-654
Davis SS, Hardy JG, Fara JW (1986) Transit of pharmaceutical dosage forms through
the small intestine. Gut 27:886-892
Ditchburn RW, Pritchard RM (1956) Avidity of the tetracyclines for the cations of
metals. Nature 177:433-434
Dobkin JF, Saha JR, Butler VPJR, Neu HC, Lindenbaum J (1983) Digoxin-inactivat-
ing bacteria: identification in human gut flora. Science 220:325-327
Dressman JB, Fleisher D, Amidon GL (1984) Physicochemical model for dose-depen-
dent drug absorption. J Pharm Sci 73(9):1274-1279
Drug Interactions in the Gastrointestinal Tract 39

Ducharme MP, Warbasse LH, Edwards DJ (1995) Disposition of intravenous and oral
cyclosporine after administration with grapefruit juice. Clin Pharmacol Ther
57(5):485-491
Faloon WW, Chodos RB (1969) Vitamin B12 absorption studies using
colchicine, neomycin and continuous 57CoB 12 administration. Gastoenterology
56:1251
Fann WE, Davis JM, Janowsky DS, Sekerke HJ, Schmidt DM (1973) Chlorpromazine:
effects of antacids on its gastrointestinal absorption. J Clin Pharmacol13:388-390
Feldman S, Carlstedt BC (1974) Effect of antacid on absorption of enteric-coated
aspirin. JAMA 227:660-661
Fermaglich J, O'Doherty SS (1972) Effect of gastric motility on levodopa. Dis Nervous
Syst 33:624-625
Fleisher D (1995) Gastrointestinal transport of peptides: experimental systems. In:
Taylor MD, Amidon GL (eds) Peptide-based drug design. American Chemical
Society, Washington DC, pp 501-525
Forrest FM, Forrest IS, Serra MT (1970) Modification of chlorpromazine metabolism
by some other drugs frequently administered to psychiatric patients. BioI Psychiatr
2:53-58
Gibbons DO, Lant AF (1975) Effects of intravenous and oral propantheline and
metoclopramide on ethanol absorption. Clin Pharmacol Ther 17(5):578-584
Gomez DY, Wacher VJ, Tomlanovich SJ, Hebert MF, Benet LZ (1995) Clin
Pharmacol Ther (in press)
Gothoni G, Neuvonen PJ, Mattila M, Hackman R (1972) Iron tetracycline interactions:
effect of time interval between the drugs. Acta Med Scand 191:409-411
Grasela TH, Schentag JJ, Sedman AJ, Wilton JH, Thomas DJ, Schulz RW, Lebsack
ME, Kinkel A W (1989) Inhibition of enoxacin absorption by antacids or
ranitidine. Antimicrob Agents Chemother 33(5):615-617
Greenblatt DJ, Allen MD, MacLaughlin DS, Harmatz JS, Shader RI (1978) Diazepam
absorption: effect of antacids and food. Clin Pharmacol Ther 24:600-609
Greenspan R, MacLean H, Milzer A, Necheles H (1951) Antacids and aureomycin.
Am J Dig Dis 18:35037
Harcourt RS, Hamburger M (1975) The effect of magnesium sulfate in lowering
tetracycline blood levels. J Lab Clin Med 50:464-468
Hayes SL, Pablo G, Radomski T, Palmer RF (1977) Ethanol and oral diazepam
absorption. N Engl J Med 296:186-189
Heinrich HC, Oppitz KH, Gabbe EE (1974) Hemmung der Eisenabsorption beim
Menschen durch Tetracyclin. Klin Wochenschr 52:493-498
Ho NFH, Merkle HP, Higuchi WI (1983) Quantitative, mechanistic and physiologically
realistic approach to the biopharmaceutical design of oral drug delivery systems.
Drug Dev Ind Pharm 9(7):1111-1184
Hoffman WS, Dyniewicz HA (1945) The effect of alumina gel upon the absorption of
vitamin A from the intestinal tract. Gastoenterology 5:512-522
Honig PK, Wortham DC, Zamani K, Conner DP, Mullin JC, Cantilena LR (1993)
Terfenadine-ketoconazole interaction. Pharmacokinetic and electrocardiographic
consequences. JAMA 269(12):1513-1518
Hsing S, Gatmaitan Z, Arias 1M (1992) The function of Gp 170, the multidrug-
resistance gene product, in the brush border of rat intestinal mucosa. Gastroenter-
ology 102:879-885
Hu M, Amidon GL (1988) Passive and carrier-mediated intestinal absorption compo-
nents of captopril. J Pharm Sci 77(12):1007-1011
Hunter J, Hirst BH, Simmons NL (1993) Drug absorption limited by P-glycoprotein-
mediated secretory drug transport in human intestinal epithelial Caco-2 cell layers.
Pharmacol Res 10(5):743-749
Hurwitz A (1971) The effects of antacids on gastrointestinal drug absorption. II. Effect
on sulfadiazine and quinine. J Pharmacol Exp Ther 179(3):485-489
Hurwitz A, Schlozman DL (1974) Effects of antacids on gastrointestinal absorption of
isoniazid in rat and man. Am Rev Resp Dis 109:41-47
40 E. LIPKA et al.

lIett KF, Tee LBG, Reeves PT, Minchin RF (1990) Metabolism of drugs and other
xenobiotics in the gut lumen and wall. Pharmacol Ther 46:67-93
Ince P, Appleton DR, Finney KJ, Moorghen M, Sunter JP, Watson AJ (1988)
Verapamil sensitises normal and neoplastic rodent intestinal tissue to the
stathmokinetic effect of vincristine in vivo. Br J Cancer 57:348-352
Iseki K, lemura A, Sato H, Sunada K, Miyazaki K, Arita T (1984) Intestinal absorption
of several f3-lactam antibiotics. V. Effect of amino f3-lactam analogues and
dipeptides on the absorption of amino f3-lactam antibiotics. J Pharm Dyn 7:768-
775
Johnson BF, O'Grady J, Bye C (1978) The influence of digoxin particle size on absorp-
tion of digoxin and the effect of propantheline and metoclopramide. Br J Clin
Pharmacol 5:465-467
Karlson J, Kuo SM, Ziemniak J, Artursson P (1993) Transport of celiprolol across
human intestinal epithelial (Caco-2) cells: mediation of secretion by multiple
transporters including P-glycoprotein. Br J Pharmacol110(3):1009-1016
Kauffman RE, Azarnoff DL (1973) Effect of colestiopol on gastrointestinal absorption
of chlorothiazide in man. Clin Pharmacol Ther 14:886-889
Kimura H, Inui K-I, Sezaki H (1985) Differences in effects on drug absorption between
dihydroxy and trihydroxy bile salts. J Pharmacobio Dyn 8:578-585
Kolars JC, Awni WM, Merion RM, Watkins PB (1991) First-pass metabolism of
cyclosporin by the gut. Lancet 338:1488-1490
Kolars JC, Schmiedlin-Ren P, Schuetz JD, Fang C, Watkins PB (1992) Identification
of rifampicin-inducible P450IIIA4 (CYP3A4) in human small bowel enterocytes.
J Clin Invest 90:1871-1878
Koster AS, Frankhiujzen-Sierevogel AC, Noordhoek J (1985a) Glucuronidation of
morphine and six beta 2-sympathomimetics in isolated rat intestinal epithelial
cells. Drug Metabol. Dispos. 13:232-238
Koster AS, Hofman GA, Frankhiujzen-Sierevogel AC, Noordhoek J (1985b)
Presystemic and systemic intestinal metabolism of fenoterol in the conscious rat.
Drug Metabol Dispos 13:464-470
Krishnamoorthy R, Mitra A (1995) Peptide metabolism by gastric, pancreatic, and
lysosomal proteinases. In: Taylor MD, Amidon GL (eds) Peptide-based drug
design. American Chemical Society, Washington DC, pp 501-525
Kuhlmann J, Zilly W, Wilke J (1981) Effects of cytostatic drugs on plasma level and
renal excretion of beta-acetyldigoxin. Clin Pharmacol Ther 30:518-527
Kuo SM, Whitby BR, Artursson P, Ziemniak JA (1994) The contribution of intestinal
secretion to the dose-dependent absorption of celiprolol. Pharmcol Res 11(5):648-
653
Kupferschmidt HHT, Ha HR, Ziegler WH, Meier PJ, Kraehenbuehl S (1995) Interac-
tion between grapefruit juice and midazolam in humans. Clin Pharmacol Ther
58(1):20-28
Lavigne J-G, Marchand C (1973) Inhibition of gastointestinal absorption of p-
aminosalicylate (PAS) in rats and humans by diphenhydramine. Clin Pharmacol
Ther 14:404-411
Lebsack ME, Nix D, Ryeerson B, Toothaker RD, Welage L, Norman AM, Schentag JJ,
Sedman AJ (1992) Effects of gastric acidity on enoxacin absorption. Clin
Pharmacol Ther 52(3):252-256
Lennernas H (1995) Does fluid flow across the intestinal mucosa affect quantitative
oral drug absorption? Is there time for a reevaluation? Pharmacol Res 12
Lennernas H, Regardh CG (1993) Dose-dependent intestinal absorption and signifi-
cant intestinal secretion (exsorption) of the beta-blocker pafenolol in the rat.
Pharmacol Res 10(5):727-731
Lennernas H, Nilsson D, Aquilonius SM, Ahrenstedt 0, Knutson L, Paalzow LK
(1993) The effects of L-Ieucine on the absorption of L-dopa, studied by regional
jejunal perfusion in man. Br J Clin Pharmacol 35(3):243-250
Levy G, Rao BK (1972) Enhanced intestinal absorption of riboflavin from sodium
alginate solution in man. J Pharm Sci 61:279-280
Drug Interactions in the Gastrointestinal Tract 41

Levy G, Tsuchiya T (1972) Effect of activated charcoal on aspirin absorption in man,

part I. Clin Pharmacol Ther 13:317-322
Lindenbaum J, Maulitz RM, Saha JR, Shea N, Butler VP (1972) Impairment of digoxin
absorption by neomycin. Clin Res 20:410
Lindenbaum J, Rund DG, Butler VP, Tse-Eng D, Saha JR (1981) Inactivation of
digoxin by the gut flora: reversal by antibiotic therapy. N Engl J Med 305:789-794
Lode H (1988) Drug interactions with quinolones. Rev Infect Dis 10 [SuppI1]:S132-
S136
Lucarotti RL, Colaizzi JL, Barry H, Poust RI (1972) Enhanced pseudoephedrine
absorption by concurrent administration of aluminium hydroxide gel in humans.
J Pharm Sci 61:903-905
Mailliard ME, Stevens BR, Mann GE (1995) Amino acid transport by small intestinal,
hepatic and pancreatic epithelia. Gastroenterology 108:888-910
Manninen V, Apajalahti A, Melin J, Karesoja M (1973) Altered absorption of
digoxin in patients given propantheline and metoclopramide. Lancet 1(800):398-
400
Muranishi S (1990) Absorption enhancers. Crit Rev Ther Drug Carrier Syst 7(1):1-33
Naggar VF, Khalil SA (1979) Effect of magnesium trisilicate on nitrofurantoin absorp-
tion. Clin Pharmacol Ther 25:857-863
Neuvonen PJ (1976) Interactions with the absorption of tetracyclines. Drugs 11:45-
54
Neuvonen PJ, Elonen E (1980) Effect of activated charcoal on absorption and elimina-
tion of phenobarbitone, carbamazepine and phenylbutazone in man. Eur J Clin
Pharmacol 17:51-57
Neuvonen PJ, Turakka H (1974) Inhibitory effect of various iron salts on the absorp-
tion of tetracycline in man. Eur J Clin Pharmacol. 7:357-360
Neuvonen PJ, Gothini G, Hackma R, af Bjorksten K (1970) Interference of iron with
the absorption of tetracyclines in man. Br Med J 4:532-534
Neuvonen PJ, Pentikainen PJ, Gothoni G (1975) Inhibition of iron absorption by
tetracycline. Br J Clin Pharmacol 2:94-96
Neuvonen PJ, Elfing SM, Elonen E (1978) Reduction of absorption of digoxin,
phenytoin and aspirin by activated charcoal in man. Eur J Clin Pharmacol13:213-
218
Ni F, Ho NFH, Fox JL, Leuenberger H, Higuchi WI (1980) Theoretical model studies
of intestinal drug absorption V. Non-steady-state fluid flow and absorption. Int J
PharmacoI5:33-47
Nimmo J, Heading RC, Tothill P, Prescott LF (1973) Pharmacological modification of
gastric emptying: effects of propantheline and metoclopramide on paracetamol
absorption. Br Med J 1:587-589
Nix DE, Watson WA, Lener ME, Frost RW, Krol G, Goldstein H, Lettieri J, Schentag
J (1989) Effects of aluminum and magnesium antacids and ranitidine on the
absorption of ciprofloxancin. Clin Pharmacol Ther 46(6):700-705
Nix DE, Wilton J, Ronald B, Distlerrath L, Williams VC, Norman A (1990) Inhibition
of norfloxacin absorption by antacids. Antimicrob Agents Chemother 34:432-435
Northcutt RG, Stiel IN, Hollifield JW, Stant EG Jr (1969) The influence of
cholestyramine on thyroxine absorption. JAMA 208(10):1857-1861
Ohnhaus EE (1980) The effect of antacid and food on the absorption of proquazone
(Biarison) in man. Int J Clin PharmacoI18:136-139
Osman MA, Welling PG (1983) Influence of prop ant he line and metoclopramide on the
bioavailability of chlorothiazide. Curr Ther Res 34(2):404-408
PaIva IP, Rytkonen U, Alatulkkila M, PaIva HLA (1972) Drug-induced malabsorption
of vitamin B 12 • Scand J Haematol 9:5-7
Parsons RL, Paddock GM (1975) Absorption of two antibacterial drugs, cephalexin
and co-trimoxazole, in malabsorption syndromes. J Antimicrob Chemother I
[Suppl]:59-67
Penttila 0, Hurme H, Neuvonen PJ (1975) Effect of zinc sulphate on the absorption of
tetracycline and doxycycline in man. Eur J Clin Pharmacol 9:131-134
42 E. LIPKA et al.

Peterson OL, Finland M, Ballou AN (1942) The effect of food and alkali on the
absorption and excretion of sulfonamide drugs after oral and duodenal adminis-
tration. Am J Med Sci 204:581-588
Pohjola-Sintonen S, Viitasalo M, Toivonen L, Neuvonen P (1993) Itraconazole pre-
vents terfenadine metabolism and increases risk of torsade de pointes ventricular
tachycardia. Eur J Clin PharmacoI45(2):191-193
Renwick AG (1982) First-pass metabolism within the lumen of the gastrointestinal
tract. In: George et al (eds) Clinical pharmacology and therapeutics: pre systemic
drug elimination. Butterworth, London, pp 3-28
Resetarits D, Bates TR (1979) Apparent dose-dependent absorption of chlorothiazide
in dogs. J Pharmacol Biopharmacol 7(5):463-470
Rietbrock N, Kuhlmann J, Leopold G (1983) Factors influencing the enteral absorption
of drugs. Inn Med 10:112-120
Rivera-Calimlim L, Dujovne CA, Morgan JP, Lasagna L, Bianchine JR (1970a) L-dopa
absorption and metabolism by the human stomach. Pharmakologist 12:269
Rivera-Calimlim L, Dujovne CA, Morgan JP, Lasagna L, Bianchine JR (1970b) L-
dopa treatment failure: explanation and correction. Br Med J 4:93
Rivera-Calimlim L, Morgan JP, Dujovne CA, Bianchine JR, Lasagna L (1970c) L-dopa
absorption and metabolism by the human stomach. J Clin Invest 49:79a
Rivera-Calimlim L, Dujovne CA, Morgan JP, Lasagna L, Bianchine JR (1971) Absorp-
tion and metabolism of L-dopa by the human stomach. Eur J Clin Invest 1:313-
320
Rivera-Calimlim L, Castaneda L, Lasagna L (1973) Effects of mode of management on
plasma chlorpromazine in psychiatric patients. Clin Pharmacol Ther 14:978-985
Robinson DS, Benjamin DM, McCormack JJ (1970) Interaction of warfarin and
nonsystemic gastrointestinal drugs. Clin Pharmacol Ther 12(3):491-495
Schreiner J, Altemeier WA (1962) Experimental study of factors inhibiting absorption
and effective therapeutic levels of declomycin. Surg Gynecol Obstet 114:9-14
Schuna A, Osman MA, Patel RB, Welling PG, Sundstrom WP (1983) Influence of food
on the bioavailability of penicillamine. J Rheumatol 10:95-97
Seed JC, Wilson CE (1950) The effect of aluminum hydroxide on serum aureomycin
concentrations after simultaneous oral administration. Bull Johns Hopkins Hosp
86:415-418
Seneca H, Peer P (1965) Enhancement of blood and urine tetracycline levels with a
chymotrypsin-tetracycline preparation. J Am Geriatr Soc 13:708-717
Sinko PJ, Hu M, Amidon GL (1987) Carrier mediated transport of amino acids, small
peptides, and their drug analogs. J Contr ReI 6:115-121
Sorby DL, Liu G (1966) Effects of adsorbents on drug absorption. II. Effect of an
antidiarrhea mixture on promazine absorption. J Pharm Sci 55:504-510
Tam YK (1993) Individual variation in first-pass metabolism. Clin Pharmacokinet
25(4):300-328
Tsuji A (1995) Intestinal absorption of j3-lactam antibiotics. In: Taylor MD, Amidon
G L (eds) Peptide-based drug design. American Chemical Society, Washington
DC, pp 501-525
Tyrer JH, Eadie MJ, Sutherland JM, Hooper WD (1970) Outbreak of anticonvulsant
intoxication in an Australian city. Br Med J 4:271-273
Veno K, Tanaka K, Tsujimura K, Morishima Y, Iwashige H, Yamazaki K, Nakata I
(1993) Impairment of cefdinir absorption by iron ion. Clin Pharmacol Ther
54(5):473-475
Wacher VJ, Wu CY, Benet LZ (1995) Mol Carcinog (submitted)
Wagner JG (1969) Aspects of pharmacokinetics and biopharmaceutics in relation to
drug activity. Am J Pharm 141:5-20
Waisbren BA, Hueckel JS (1950) Reduced absorption of aureomycin caused by alu-
minium hydroxide gel (Amphojel). Proc Soc Exp BioI Med 73:73-74
Walter E, Kissel T, Amidon GL The intestinal peptide carrier: a potential transport
system for small peptide derivated drugs. Adv Drug Deliv Rev (accepted)
Drug Interactions in the Gastrointestinal Tract 43

Watkins PB, Wrighton SA, Schuetz EG, Molowa DT, Guzelian PS (1987) Identifica-
tion of glucocorticoid-inducible cytochromes P-450 in the intestinal mucosa of rats
and man. J Clin Invest 80:1029-1036
Weinberg ED (1957) The mutual effects of antimicrobial compounds and metallic
cations. Bacteriol Rev 21:46-68
Welling PG (1984) Interactions affecting drug absorption. Clin Pharmacol 9:404-434
Wetterich U, Mutschler E, Spahn-Langguth H, Langguth P (1995) Evidence for intes-
tinal secretion of the f3-adrenoceptor antagonist talinolol: data from humans and
studies with Caco-2 cells. Naunyn Schmiedenbergs Arch Pharmacol 351
[Suppl]:R1
Whitehouse LW, Paul CJ, Coldwell BB (1975) Effects of ethanol on diazepam distribu-
tion in rats. Res Commun Chern Pathol Pharmacol12:221-242
Williamson RM, Oxender DL (1995) Molecular biology of amino acid, peptide and
oligopeptide transport. In: Taylor MD, Amidon GL (eds) Peptide-based drug
design. American Chemical Society, Washington DC, pp 501-525
Yamaguchi T, Ikeda C, Sekine Y (1986a) Intestinal absorption of a f3-adrenergic
blocking agent nadolol. I. Comparison of absorption behavior of nadolol with
those of other f3-blocking agents in rats. Chern Pharm Bull 34(8):3362-3369
Yamaguchi T, Ikeda C, Sekine Y (1986b) Intestinal absorption of a f3-adrenergic
blocking agent nadolol. II. Mechanism of the inhibitory effect on the intestinal
absorption of nadolol by sodium cholate in rats. Chern Pharm Bull 34(9):3836-
3843
Yamaguchi T, Ikeda C, Sekine Y (1986c) Intestinal absorption of a f3-adrenergic
blocking agent nadolol. III. Nuclear magnetic resonance spectroscopic study on
nadolol-sodium cholate micellar complex and intestinal absorption of nadolol
derivatives in rats. Chern Pharm Bull 34(10):4259-4264
Yee S, Amidon GL (1995) Oral absorption of angiotensin-converting enzyme inhibi-
tors and peptide prodrugs. In: Taylor MD, Amidon GL (eds) Peptide-based drug
design. American Chemical Society, Washington DC, pp 501-525
CHAPTER 3
Drug-Food Interactions
Affecting Drug Absorption
P.G. WELLING

A. Introduction
The intent of this book is to present current information on drug interactions
influencing drug absorption, distribution, elimination, and activity. Consistent
with this, the original plan of the authors concerned was to present a single
chapter on interactions affecting drug absorption which would include drug-
drug interactions and also drug-food interactions. After examining the quan-
tity of material that has been published on these two closely related but
distinct topics, it was agreed that drug-drug interactions and drug-food inter-
actions affecting drug absorption would be better presented in two separate
chapters.
This chapter then, is devoted exclusively to drug-food interactions affect-
ing drug absorption. No attempt has been made to address the related topic of
the effect of drugs on food and nutrient absorption. Material on this topic has
been reviewed elsewhere (BASU 1988; ROE 1989; TROVATO et al. 1991).
While a number of studies on drug-food interactions were reported
earlier, the subject did not achieve full recognition in scientific and regulatory
circles until the first major review on this topic (WELLING 1977). Since that
time the number of studies examining various aspects of drug-food interac-
tions affecting drug absorption has increased dramatically and many of the
results obtained from these studies have been quite spectacular, not only in the
extent of some interactions but also in their unexpected and totally unpredict-
able nature. The topic has been reviewed extensively, not only in terms of
interactions of particular drugs or families of drugs (WELLING 1989, 1993;
PFEIFER 1993; SORGEL and KINZIG 1993; WELLING and TSE 1982; WILLIAMS et al.
1993), but also relating to clinical consequences and the management of
drug-food interactions in the clinical environment (LASSWELL and LORECK
1992; NEUVONEN and KMSTO 1989). The number of reviews on this topic
reflects increased awareness in scientific, regulatory, and clinical circles of its
importance.
The original goal of many investigators in this area, including the writer,
was to establish mechanisms involved in drug-food interactions and also to use
data obtained from a variety of situations to establish rules or guidelines to
predict the nature and extent of interactions in terms of circulating drug or
metabolites. As with so many scientific goals this has proven to be elusive.
46 P.G. WELLING

After almost 20 years of research it is still difficult, and many times impossible,
to predict the nature of drug-food interactions and the literature is replete with
surprises, and unpredictable results.
Based on current knowledge, the unpredictable nature of various inter-
actions is probably related, not only to the complex environment in which
interactions occur, but also to the many different ways in which they have been
examined. Some of them are summarized here.

I. Food
The type and size of meal may have a marked effect on the nature of a drug-
food interaction. One might not expect that a meal prepared in England, in
India, or in Japan would act similarly, nor might a breakfast compared to an
evening meal. Similarly liquid meals, which are so often used in an attempt to
obtain mechanistic information, might have a totally different effect on drug
absorption than solid meals, which are usually more clinically relevant. Also
the time interval between eating and medication will influence the nature and
extent of a drug-food interaction.

II. Dosage Form

In addition to the interaction between food components or their pharmaco-
logical sequelae on the gastrointestinal (GI) tract and drug molecules, must be
considered the effect of the formulation. Much of the data summarized in
previous reviews, and also to be reviewed in this chapter, have demonstrated
the dramatic effect that formulation may have on drug-food interactions. This
has been particularly important and controversial with controlled release
products. Studies on these formulations have given rise to the unfortunate
term "dose dumping," often sadly misinterpreted. The preponderance of evi-
dence shows that the more disperse a formulation, the less the effect of drug-
food interactions. Thus, drugs administered in solution are likely to be much
less affected by food than drugs formulated in a compressed tablet. Such is the
extent of the literature on this topic that this review is restricted for the most
part to material published since 1990. Material published before 1990 can be
readily obtained from previous reviews or from literature cited here.

B. Influence of Food on the Gastrointestinal Tract

The possible effects that ingested food may have on GI physiology are summa-
rized in Table 1. This subject has recently been reviewed by OOSTERHUIS and
JONKMAN (1993). In the fasting state, gastric motility passes through cycles of
migrating motor complexes (MMC). Each cycle lasts 2-3h and comprises four
phases, of which phase 3, the "housekeeper wave," is the strongest. Non-
nutrient liquids pass through the stomach throughout the MMC but solids are
Drug-Food Interactions Affecting Drug Absorption 47

Table 1. Possible effects of ingested food on gastrointestinal tract physiology. (From

WELLING 1993)

Physiological function Effect of food Possible effect on drug absorption

Stomach emptying rate Decreased rate with Absorption generally delayed,

solid meals, fats, may be reduced with unstable
high temperature, compounds, may be increased due
acids, solutions of to drug dissolution in stomach.
high osmolarity. Absorption increased with large
Increased rate fluid volumes
with large fluid
volumes
Intestinal motility Increased Faster dissolution and decreased
diffusional path promotes
absorption. Shorter transit time
may inhibit absorption
Splanchnic blood flow Generally increased, Absorption increased with faster
but may be blood flow. Variable effects on
decreased by first-pass metabolism, depending on
ingestion of glucose drug characteristics
Bile secretion Increased Absorption may increase due to
faster dissolution or may decrease
due to complexation
Acid secretion Increased Increased absorption of basic drugs
provided they are acid stable.
Decreased absorption of acid-labile
compounds
Enzyme secretion Increased Increased or decreased absorption
depending on drug characteristics
Active absorption Decreased Active drug absorption reduced by
progress competitive inhibition

moved from the stomach into the intestine mainly during phase 3. Depending
on when a solid meal is ingested relative to the MMe, the gastric residence
time may vary from a few minutes to 2-3 h (EWE et al. 1989). Ingested food
changes gastric motility to a postprandial pattern during which time gastric
residence time is increased. Gastric residence time is increased by large, par-
ticularly solid meals and by chyme of low pH, high osmolality, and high fat
content. Residence time is also influenced by hot meals.
While solid foods tend to delay gastric emptying, non-nutrient liquid
meals, for example, a fluid volume ingested with medication, may have the
opposite effect. As described previously in the review by WELLING (1977), the
stomach appears to empty liquid meals into the duodenum in apparent first-
order fashion. Distension of the stomach is the only natural stimulus known to
increase stomach emptying and the observed faster emptying rate with in-
creasing fluid volume can be rationalized in terms of varying tension at recep-
tors in the stomach wall (HOPKINS 1966). The presence of ingested food also
promotes gastric secretion of hydrochloric acid. Once food passes from the
48 P.G. WELLING

stomach into the small intestine it has a stimulating effect on intestinal motil-
ity, on digestive enzyme secretion, and also, particularly in the case of fatty
meals, on bile secretion.
The influences of altered gastric and intestinal motility, the one decreased
and the other increased, and increasing GI secretions with ingested food,
might be expected to have a number of possible effects on drug absorption. All
of these have been proposed, in one way or another, to explain observed
results of drug-food interactions. Delayed gastric emptying will likely delay
absorption of drugs that are absorbed predominantly from the small intestine,
but not of drugs that are efficiently absorbed from the stomach. It may delay
the absorption of acidic compounds or drugs in enteric-coated formulations by
delaying drug transit from the acidic gastric contents to the relatively alkaline
region of the small intestine. On the other hand, delayed gastric emptying
might increase systemic availability of compounds that are slowly soluble at
acidic gastric pH by permitting more material to dissolve in the stomach
before passing into the small intestine. Compounds that are unstable in acidic
pH are likely to be degraded as a result of prolonged residence in the stomach
regardless of the extent to which pH may be lowered, or raised, by the intake
of food. The latter point is important because, while solid foods stimulate
acidic secretion, the buffering effect of different types of meals may minimize
any change in pH and could even give rise to a transient increase in gastric pH.
Increased intestinal motility in the fed state may have the effect of increas-
ing dissolution of solid particles and could also decrease the mean diffusional
path of dissolved molecules to the intestinal mucosa. On the other hand, it
could increase the transit rate of molecules from the efficient absorption
region of the small intestine to the less efficient region of the large intestine,
thus negatively impacting absorption efficiency.
The effect of fluid volume on GI physiology adds another dimension to the
complexity of drug-food interactions. It has already been noted that large
volumes of non-nutrient fluids empty from the stomach at a faster rate than
small volumes. The large volume itself tends to accelerate dissolution and also
to accelerate transfer of both dissolved and undissolved drug into the small
intestine. The presence of large fluid volumes in the intestine may continue to
increase drug absorption by providing a more liquid environment and also by
a solvent drag effect from the serosal to the mucosal side of the intestinal
membrane. On the other hand, and from a physicochemical perspective, a
large fluid volume might be expected to delay or reduce absorption due to a
reduced serosal-mucosal drug concentration gradient. In the original review
on this subject (WELLING 1977) it was noted that the positive effects of large
fluid volumes appear often to outweigh the negative effects, and more
examples of increased drug absorption when they are administered with large
fluid volumes have appeared in the literature to support this argument. The
role of gastric motility in drug absorption, with particular reference to the
definition of "fasting" and "non-fasting" drug administration, has recently
been reviewed by WALTER-SACK (1992).
Drug-Food Interactions Affecting Drug Absorption 49

In addition to its effects on GI motility and secretion, ingested food may

also affect splanchnic blood flow. However, the degree and direction of change
may vary with the types of food. For example, a high-protein meal has been
shown to cause a 35% increase in splanchnic blood flow, while a liquid glucose
meal caused an 8% decrease. In most cases after solid meals, one would expect
splanchnic blood flow to increase (McLEAN et al. 1978) and this should pro-
mote drug absorption via the splanchnic circulation. How much drug subse-
quently enters the systemic circulation is a function also of presystemic
clearance and how this is affected by food-related changes in splanchnic circu-
lation. The most frequently documented food effect has been that of increased
drug systemic availability associated with reduced presystemic clearance. This
has been reported for propranolol, hydralazine, metoprolol, and propafenone
(MELANDER and McLEAN 1983; AXELSON et al. 1987). A number of studies
have investigated this phenomenon and have shown that a transient postpran-
dial increase in splanchnic blood flow is associated with a decrease in
presystemic hepatic extraction; the magnitude of this effect correlates with
protein content of a meal (LIED HOLM et al. 1990).
Apart from its indirect and variable inference on pre systemic hepatic
clearance associated with altered splanchnic blood flow, food may also exert a
direct influence on intrinsic drug metabolism, both presystemic and systemic.
For example, it has been known for some time that oxidative metabolism may
be increased by high-protein diets or diets high in cruciferous vegetables. A
more recent example of a direct interaction is acute inhibition by grapefruit
juice on the metabolism of dihydropyridine calcium channel blockers. Sys-
temic availability of orally administered felodipine was increased more than
twofold by co administered grapefruit juice (BAILEY et al. 1991). A similar but
less pronounced effect was observed with nifedipine. A postulated mechanism
for this effect is linked to possible inhibition of cytochrome P450 IIIA4 by
flavonoids present in grapefruit juice, but this has yet to be confirmed.

c. Direct Effect of Food on Drug Absorption

In addition to the indirect effects that food may exert by altering GI physiol-
ogy, food may also affect drug absorption by direct interaction. Some of the
direct effects of food affecting drug absorption are summarized in Table 2.
Food may act as a physical barrier, inhibiting drug dissolution and preventing
access of dissolved drug to the mucosal surface of the GI tract. Fluid volume
may present a direct physical interaction by its effect on drug or formulation
dissolution rate and solubility together with the opposing effects of solvent
gradient and transmucosal solvent drag. Specific ions or other substances in
food may interact with drug, resulting in reduced drug absorption. Examples
are chelation of tetracycline and penicillamine by metal ions and also complex
formation by proteins. For the small number of drugs that are actively ab-
sorbed, although this number is likely to increase with the introduction of
50 P.G. WELLING

Table 2. Direct drug-food interactions that may influence drug absorption. (From
WELLING 1993)

Interaction Possible effect on drug absorption

Food and food Absorption decreased due to chelation, absorption, adsorption,
components physical blockade, but may be increased due to increased
solubility with some diets. Variable effects due to pH changes
Fluid volume Absorption rate may decrease with large fluid volumes due to
reduced concentration gradient, but absorption efficiency may
be increased due to faster dissolution, osmotic effect, and
exposure of drug molecules to greater GI surface area

more protein-derived drugs, direct competition for active carriers may occur
between food protein and protein fragments and drug molecules, inevitably
giving rise to decreased drug systemic availability.
Given the multifactorial nature of drug-food interactions, and the large
number of variables that need to be addressed in one way or another in studies
of this type, it is not surprising that variable and often conflicting results are
obtained from different laboratories. In order to better understand and inter-
pret these types of studies it is essential that study conditions be carefully
controlled and completely described. The writer makes no plea for uniformity
in studies, on the contrary. Only by studying interactions under widely varying
conditions will it be possible to completely understand them, together with
their mechanistic and clinical implications. Nonetheless, complete study de-
tails must be reported just as in the recording of a chemical or physical
experiment. Different studies have different goals. For example administering
a drug together with the well-known English breakfast addresses a clear
clinical question, relevant at least to those who indulge in extravagances, but
does not attempt to address the mechanistic question. Administering a drug
together with a high-protein fixed-volume liquid meal, on the other hand,
attempts to address the mechanistic question but tends to be clinically irrel-
evant. Which study is the most important depends on one's perspective. How-
ever, both study types are necessary in order to completely understand the
phenomenon. It is interesting to note that the extent to which drug absorption
may be altered by the presence of food, while in some cases may be slight or
even negligible, may in other cases be far greater than would be permitted by
any regulatory agency if it were considering interchangeability of generic
products. The problem of drug-food interactions has been exacerbated by the
introduction of controlled release oral products. These devices may serve a
useful clinical role in generating desirable plasma drug profiles, possibly lead-
ing to better control of pathological conditions, and also increasing patient
compliance. However, they also present a greater quantity of drug to the
patient per single dosing unit than conventional dosage forms, and are de-
signed to deliver drug at a controlled rate over a prolonged period. As these
dosage forms are also given less frequently than conventional formulations, a
Drug-Food Interactions Affecting Drug Absorption 51

marked effect by administered food on systemic availability may have a seri-

ous and prolonged consequence on circulating drug levels.
The writer finds it convenient in reviews of this type to divide drugs or
drug formulations into those whose absorption is reduced, delayed, increased,
or not affected by food. The great majority of cases fall into one or more of
these four categories. However, a small number of studies have reported
accelerated but not increased absorption, and this category is new. The tables
summarizing here the results obtained in these various categories provide as
much information as can be practically included, describing the formulation
and also study details where they are available. Lack of information in some
cases may be due to the brevity of information in published abstracts, rather
than full papers. The format of the tables is similar to that used in the original
review on this topic (WELLING 1977). Although most of the reported drug
studies during the review period are, to the writer's knowledge, included, the
treatment is intended to be representative rather than exhaustive.

D. Interactions Causing Reduced Drug Absorption

Studies that have demonstrated reduced drug levels and/or urinary excretion
of drugs are summarized in Table 3. In this and the following tables, com-
pounds are listed alphabetically.
Alendronate sodium is a potent biophosphonate under investigation
for treatment of osteoporosis. The systemic bioavailability of alendronate
averages approximately 0.75% of the administered dose, with considerable
intra-, and intersubject variation (GERTZ et al. 1993). Absorption is further
reduced by approximately 40% when alendronate is taken 1-2h before
food, and by as much as 90% when taken up to 2h postprandially. Although
the moderate decrease in absorption when alendronate is taken before
meals may have marginal clinical significance, the more dramatic reduction
with postprandial dosage may well be significant. This has led to a recommen-
dation that alendronate be taken after overnight fast, at least 30min
before food. No explanation was offered for reduced absorption of the already
poorly absorbed alendronate, but it was noted that alendronate forms in-
soluble complexes with multivalent cations such as Ca2+ and Mg2+, which may
contribute.
Dramatic reduction has been reported in serum levels of the cholinest-
erase inhibitor ambenonium chloride by food in male and female patients with
myasthenia gravis (OHTSUBO et al. 1992). Following ingestion of one or one-
half a Mytelase (ambenonium chloride) lO-mg tablet, peak serum concentra-
tion ( ern •x) and the area under the serum concentration profile from zero to 3 h
after dosing [AUC(0-3)] were both reduced by 70% after a conventional
breakfast compared to the fasting state, while the time of peak concentration
(trn.x) was moderately increased. Mean serum-time curves following single
10- and 5-mg doses of ambenonium chloride are shown in Figs. 1 and 2.
Table 3. Drug-food interactions resulting in reduced drug absorption
Drug Dosage form Dose Dosage Subjects Food details Fluid volume Time interval Sampling Result of food Reference
regimen duration administration

Alendronate SDb 0.5-1 h before, Absorption reduced 40% GERTZ et a1.

sodium with, or up to when dosed 0.5-1 h (1993)
2 h after meal before food, and by
80%-90% after food
Ambenonium Tablet 5 or 10 mg SD 13 patients with Conventional 150ml water 30 min after meal 3 h, nerum Serum C~ and AUC OHTSUBO et a1.
chloride myasthenia gravis breakfast values reduced by over (1992)
6M,7F 70%, Tmax moderately
increased
Atenolol Tablet, capsule 50mg 8 healthy subjects No food 250ml water NR' 24 h plasma C_ reduced 28%, AUC BARNWELL et a1.
with bile acids (6M,2F) reduced 30% by bile (1993)
acids
Azithromycin 52 % decrease in Cmax, DREW and
43% decrease in AVe GALLIS (1992)
due to large meal
Cefprozil 7.5 mglkg 6 10- to 14-year-old Before/after meal 8h serum Cmu reduced 22%, urine NAKAMURA and
children and urine excretion by 16% IWAI (1992)
Ceftibuten Capsule 250mg SD Enteruedd 200ml water - 10 h plasma Urinary ceftibuten ISEKI et a1.
Hepan ED 11 h urine excretion markedly (1994)
decreased by both
elemental liquid diets
2-Chloro-2'- Solution 0.24mglkg SD 7 male patients with Standard breakfastd 24h plasma Cmn reduced 40%, AVe ALBERTION!
deoxyadenosine leukemia reduced 9% et a1. (1993)
2-Choro-2'- SD 4 patients with Plasma Cmax reduced 40%, ALBERTIONI
deoxyadenosine chronic lymphatic systemic availability 11 %. et al. (1993)
leukemia Tmax increased by 0.8 h
Cicaptrost Tablet 5,7.5,lOmg MD' 8 healthy male 1 h after meals Heavier meals (lunch, BELCH et a1.
subjects 9 a.m., 2 p.m., dinner) may attenuate (1993)
7 p.m., 9 a_m. antiplatelet effect
Ciprofloxacin Tablet 500mg SD 7 healthy volunteers Milk, yoghurtd 300ml water Immediately 24h plasma Bioavailability reduced NEUVONEN
3M,4F before milk, water, and urine 30% by milk, 36% by et al. (1991)
or yoghurt yoghurt
Didanosine Chewable 300mg SD 10 males Standard breakfastd Ihor30min 12h plasma No effect when drug KNUPP et al.
tablet seropositive for before, 1 h or and urine taken before meal. emu (1993)
HIV but free of 2h after meal reduced 60% and AUC
AIDS symptoms 55 % when dose taken
1 or 2 h after meal
Didanosine Chewable 375mg SD 8 healthy male Standard breakfastd 120ml water 20 min after 12h plasma C_ reduced 54%, AUC SHYU et a1.
tablet subjects starting meal and urine 47%, and urinary (1991)
excretion by 50%
Dideoxycytidine Tablet 1.5 mg SD 20 HIV positive Standard breakfast With meal Cmax reduced 35%, AVe NAZARENO et a1.
patients reduced 14% (1992)
Doxazosin Tablet Img SD 12 hypertensive Standard breakfastd 200ml water With meal 48h plasma Moderate reduction in CONWAyet al.
subjects Cmn and AUC, also (1993)
moderate reduction in
effect on blood pressure
Flecainide 2Smgq6h 1- to 6-day-old male Milkd Single Serum concentrations RUSSEL and
infant samples reduced by milk feeds MARTIN (1989)
Hydralazine Solution SOmg SD 8 healthy subjects Standard breakfastd, 100ml water With meal 6h plasma Cmu substantialIy reduced, SEMPLE et al.
4M,4F liquid nutrient AVC moderately (1991)
bolus or infusion reduced by breakfast and
bolus nutrient
Levodopa, Controlled Variable MD 12 patients (7M,SF) Standard luncheond Together with meal 4 h plasma Lower plasma levodopa Roos et al.
carbidopa release with idiopathic and carbidopa levels with (1993)
Parkinson's disease meal, but 3-0-
methyldopa levels
increased
Metformin Tablet 8S0mg SD 24 healthy male After meal C mu reduced 35%, AVe BROOKES et al.
subjects reduced 20% (1991)
Methotrexate 2.7-18 mg/m2 SD 14 juvenile Breakfast of choice Immediately after 6h plasma C max reduced 38%, DUPUIS et a1.
rheumatoid arthritis meal AVC reduced lS% (1992)
patients 4M, 10F
MK-679 Tablet SOOmg SD Healthy male 12h plasma Plasma AVC reduced YOGENDRAN
subjects 40% et a!. (1992)
Naproxen Controlled 7S0mg SD 12 healthy subjects, Standard breakfast d 200ml Immediately after 48h plasma C max reduced 14%, AVe PALAZZINI et al
release tablet 7M,SF orange juice meal reduced 10% (1992)
or water
Navelbine Capsule 80 mg/m2 SD 13 patients, 4M, 9F, Standard meal C_ and AVC reduced LUCAS et aI.
with refractory 10% (1993)
Nitrendipine Tablet, whole Variable MD 13 juvenile patents ';1Oml water - 6 or 8 h Cm.. and AVC WELLS and
or fragmented 0.5-17 years plasma significantly reduced and SINAIKO (1991)
Tmax significantly
increased by food
Norftoxacin Tablet 200mg SD 7 healthy subjects Milk, yoghurtd 300ml water Dosed immediately 24 h plasma Serum C max reduced 50% KIVISTO et a1.
4M,3F before milk, water, and urine by milk and yoghurt. (1992)
or yoghurt AVC(0-24) reduced 48%
by milk and S8% by
yoghurt
Paracetamol Solution in 20mg/kg SD 10 healthy Thai Daily dietsd 400ml water After overnight 24h plasma Cmax reduced 25%, AVe PRESCOIT et a1.
coca cola subjects on fast and urine 12 %, and urinary (1993)
vegetarian diet, 10 recovery by 16%
on normal diet in vegetarians
Phenytoin Suspension 360mg/day MD 1 67-year-old male Liquid food Single Serum phenytoin levels TAYLOR et al.
patient concentrate samples reduced by nasogastric (1993)
inquid food
Pravastatin Tablet 20mg MD 74 Standard 600- with or 1 h before 12 h serum C_ reduced 49%, AVC PAN et a!.
hyperchosterolemic to 6S0-cal. meal meal, 4 weeks reduced 31 % (1993)
male patients
RO 42-5892 Capsule 600mg SD 22 subjects Standard breakfast Immediately after 24h plasma emax reduced 59%-63% WEBER et al.
or 1 h after meal when dosed immediately (1993)
after meal, and 10%-61 %
when dosed 1 h after
meal
Ruftoxacin 800mg SD 10 healthy male 96 h plasma Cmax reduced 17%, AVe SEGRE et a1.
subjects and urine (0-96) reduced 33%. (1992)
Converse increase in
Ae(0-96)
Table 3. Continued
Orug Dosage form Dose Dosage Subjects Food details Fluid volume Time interval Sampling Result of food Reference
regimen duration administration

SK & F 106203 Tablet 150mg SO 12 healthy male Plasma levels reduced NICHOLS et at.
subjects and delayed (1992)
Sotalol 160mg SO 18 healthy subjects Standard meal 36h plasma Bioavailability reduced HANYOK (1993)
approximately 20%
Sulpiride Film-coated 100mg SO 3 healthy male Standard light, 100ml water With meal 48 h, urine Cumulative urinary SHINKUMA et al.
tablet, solution subjects medium and heavy excretion reduced by 30% (1990)
mealsd following both dosage
forms
Tacrine Capsule 40mg SO 24 healthy subjects Standard breakfastd 8 fluid 15 min and 2h 14 h plasma Cmax reduced 45% and WELTY et al.
ounces water after starting meal AUe 23% when taken (1994)
15 min after starting meal.
Cmax reduced 35 % and
AU C 21 % when taken
2 h after starting meal
Tetracycline Capsule 250mg SO 9 healthy subjects Standard meal and 200mi water Immediately after 72 h urine Urine excretion reduced COOK et a1.
high-calcium meal 40% by standard meal and (1993)
Mexican meald 55% by Mexican meal
Verapamil SR tablet 240mg SO 9 halthy male Immediately after 30h serum 48 % decrease in C max and HOON et al.
subjects meal 30% decrease in AUe, (1992)
also altered electro
cardiographic effects
Zidovudine 100mg, SO 8 male 2500 kJ breakfast, 200ml water Mid-meal 6h plasma Cma~ and AUe reduced LOITERER et al
250mg AIDS patients fat content 40 g 64% and 29%, (1991)
respectively
Zidovudine Capsule 250mg SO 13 AIDS patients, High-fat liquid meal 600ml water Immediately 4 h plasma C max reduced 50%, UNADKAT et al.
IOM,3F given as milk shaked after meal and urine T max increased (1990)
approximately threefold
Zidovudine Capsule 100 mg, SO 25 patients, 24M 3F, Standard breakfastd 100ml water Immediately after 6 h plasma, C max and AUe reduced RUHNKE et al
250mg with HIV infection meal 24 h urine 35% at 100-mg dose, (1993)
Cmax reduced 73% and
AUC 14% at 250-mg dose
Zidovudine 200mg SO HIV-I-positive After meal 4h serum Serum C max reduced UEDA et aL
patients 50% and AUC 38%. (1993)
Urinary recovery reduced
by 19%. Serum levels of
glucuronide also reduced
but urinary recovery
unchanged
882C 200mg SO 15 healthy male Standard breakfast em.. reduced 16%, AUC PECK et aL
subjects reduced 10%. Food (1993)
reduced variability in
AUC

a Data not available.

b Single dose.
c Not relevant.
d ~.1eal described.
e Multiple doses.
Drug-Food Interactions Affecting Drug Absorption 55

15
E *

-1
~
E
*

1/ -l. . j
~
"'C
.;:
0
10
::c<.I
E
:s
'c0
c
~
..Q 5
E
« I
I
E I
...
:s I
~
til
I
I
0
0 60 120 180
Time (min)

Fig. 1. Mean serum versus time concentrations (±SE) of ambenonium chloride after
oral administration of lOmg ambenonium chloride to six myasthenia gravis patients
underfasting (0) and nonfasting (e) conditions. *P, <0.05; **P, <0.01. (From OHTSUBO
et al. 1992)

Significant differences were observed between fed and fasted serum concen-
trations at 1.5, 2, and 3 h (P < 0.05) following the lO-mg dose. Following the
5-mg dose, significant differences were observed at 1 (P < 0.005),1.5,2, and 3 h
(P < 0.05).
Ambenonium chloride is a quaternary ammonium compound and is un-
likely to be well absorbed, even in the fasting state. Further reduction of
absorption due to food could be due to a number of factors, but these were not
examined in this study. Despite the rather dramatic food effect, relationships
between serum drug levels and relative clinical effect (changes in muscle
strength) were variable and inconclusive. It is nonetheless recommended that
ambenonium chloride dosage be adjusted during non-fasting treatments when
the adjusting the optimum regiment for patents with myasthenia gravis.
Previous studies with atenolol have shown that absorption is reduced by
food (MELANDER et al. 1979). In this respect the very hydrophilic atenolol
molecule differs from the lipophilic compounds of this class, metroprolol and
propranolol, both of which undergo extensive pre systemic metabolism and
whose absorption is increased by food. The mechanism by which food de-
creases atenolol absorption was addressed in a study conducted in healthy
volunteers who received single 50-mg doses of atenolol as a commercial tablet,
or as capsules containing bile acids (BARNWELL et al. 1993). In this study the
bile acid formula reduced mean atenolol plasma e max by 28% and AVe by
56 P.G. WELLING

15

~
5
QJ
~
.§ 10
::cu **
E
::::I
·2 *

l-L-t~
o
c:
~
E
5 *
<C

.
E
::::I
QJ
III I
/ -----~
I,A--~--~--~-- __~
O~~---,------~----~
o 60 120 180
Time (min)
Fig. 2. Mean serum versus time concentrations (±SE) of ambenonium chloride after
oral administration of 5 mg ambenonium chloride to seven myasthenia gravis patients
under fasting (0) and nonfasting (.) conditions. *P, <0.05; **P, <0.01. (From OHTSUBO
et al. 1992)

30%. As both formulations were administered in the fasting state it is claimed

that bile acids reduce systemic availability of atenolol to a similar or greater
extent to that previously reported by ingested food, although the mechanism
of this effect was not postulated. Previous studies have shown that bile acids
can have the effect of increasing (S.K. COLE et al. 1992) or decreasing
(YAMAGUCHI et al. 1986) drug absorption. The use of different formulations in
the present study, i.e., commercial tablets and contemporaneously prepared
capsules containing bile acids with atenolol, may also have contributed to the
observed results.
An extensive food effect reducing drug absorption is implied by summary
data contained in a review article on azithromycin (DREW and GALLIS 1992).
Following a 500-mg oral dose, systemic availability of this azalide antibacterial
agent is approximately 30%. However, coadministration with a large meal
may further reduce absorption by up to 50%. The mechanism of this interac-
tion is unknown. This basic, highly lipid soluble and poorly water soluble
compound is degraded by acid-catalyzed hydrolysis of the ether bond to the
neutral cladinose sugar (FIESE and STEFFEN 1990). Regardless of the possible
mechanisms involved, the manufacturers recommend that azithromycin be
taken either 1h before or 2h after meals. Two studies have demonstrated
moderately reduced absorption of new cephalosporins, one by food in children
(NAKAMURA and IWAI 1992a), the other by an elemental diet (ISEKI et al. 1994).
In the first study, conducted in six children, administration after a meal caused
Drug-Food Interactions Affecting Drug Absorption 57

a 22% reduction in mean Cmax values and a 16% reduction in urinary excretion
of cefprozil. In the second study, simultaneous administration of an elemental
diet Enterued, composed of oligopeptide (egg white hydrolysate), an elemen-
tal diet Hepan ED composed of amino acids, and also a mineral solution
containing neither pep tides nor vitamins had a similar effect on ceftibuten.
Plasma level data in rats, and also in situ rat jejunal loop studies, confirmed the
effect of these test meals on ceftibuten absorption. These results contrasted
with those from cephalexin, which was unaffected by Enterued formula, show-
ing that the inhibitory effects of Enterued on absorption were not due to
inhibition of a peptide transport system.
Administration of a standard breakfast moderately reduced the systemic
availability of oral 2-chloro-2'-deoxyadenosine (CdA) in male patients with
leukemia (ALBERTIONI et al. 1993). The Cmax of CdA in plasma was reduced by
40% while the AVC was reduced by 9% by food. In this study, pretreatment
with omeprazole did not significantly influence CdA bioavailability, or
interindividual variability in the fasting state. The absolute bioavailability of
CdA is only 50% when administered to fasting patients with and without
omeprazole. CdA has a low pKa such that most of the drug will be ionized in
the gut, hindering absorption. The drug may also undergo first-pass clearance
in the gut wall and/or the liver. Reduced systemic availability in the presence
of food may be related to slower absorption, resulting in a greater proportion
of absorbed drug undergoing presystemic metabolism. Reduced absorption
due to food may be a possible explanation for reduced antiplatelet activity of
cicaprost (BELCH et al. 1993). Healthy volunteers received varying doses of
cicaprost at 9 a.m., 2 p.m., 7 p.m., and again at 9 a.m. the next day. Antiplatelet
activity was dose related but the effect was attenuated following the 2 p.m. and
7 p.m. doses, which were taken after heavier meals than the 9 a.m. doses. An
alternative explanation is that tachyphylaxis may occur during the second (2
p.m.) and a third (7 p.m.) dose, causing reduced efficacy, while the 14-h gap
between the 7 p.m. dose and the next 9 a.m. dose may have been sufficient for
platelet sensitivity to return to normal. More studies are necessary to differen-
tiate these possible mechanisms. Although neither a standard breakfast nor a
high-fat, high-calcium breakfast altered ciprofloxacin absorption (FROST et al.
1989), absorption was markedly impaired by co administered milk and yoghurt
(NEUVONEN et al. 1991). Systemic availability was reduced 35% by
co administered 30ml milk and 36% by 300ml yoghurt compared to 300ml
water. Ciprofloxacin together with other fluoroquinolones is known to bind to
heavy metals including calcium to form an insoluble chelate. The reason for
the different results from the two studies is claimed to be due to calcium in
milk and yoghurt being in liquid form compared to the previous study in which
most of the high calcium present was in solid form, for example, in cheese. This
suggests, that, for at least some dairy products, ciprofloxacin is affected to a
similar extent as tetracycline derivatives (MATTILA et al. 1972). Plasma levels of
ciprofloxacin taken with milk, yoghurt, and water are shown in Fig. 3.
58 P.G. WELLING

The timing of food ingestion relative to dosing had a marked effect on the
absorption of the purine nucleoside analogue didanosine (KNUPP et a1. 1993).
This study, conducted in ten men seropositive for human immunodeficiency
virus (HIV) but free of AIDS symptoms, examined the absorption of
didanosine following overnight fast, 13 min before, 1 h before, 1 h after, and 2 h
after a standard high-fat, high-calorie breakfast. There were no significant
differences among the fasting, 30min, and 1 h before meal doses. However,
Cmax , AUC(O-oo), and urinary excretion were significantly reduced following
the two postprandial doses. The mean profiles following the five treatments
are summarized on a logarithmic scale in Fig. 4.
Decreased bioavailability after postprandial administration may have
been due to prolonged gastric retention leading to increased degradation of
the acid-labile didanosine, possible interference with active transport, or direct
interaction of didanosine with one or more meal components. In a separate
study using a chewable tablet form of didanosine in male subjects seropositive
for HIV, the bioavailability of didanosine was reduced approximately twofold
when it was administered 5 min after a substantial, standardized breakfast

o~~~~--~--~------~----~------~--~
o 1 2 3 4 6 8 10 24
Time (hr)

Fig. 3. Mean plasma versus time concentrations (±SE) of ciprofioxacin in seven sub-
jects following a single 500-mg oral dose of ciprofioxacin with 300ml milk (e), yoghurt
(.), or water (0). (From NEUVONEN et a1. 1991)
Drug-Food Interactions Affecting Drug Absorption 59

2000,-------------------------------~
-0- Fasting
-0- 30 min Before
-b- 1 hr Before
...... 1 hr After
-(too 2 hr After

~5
Q,j
c
'iii
g 200
:a'"
"C

'~E"
c:::

2~--~----_+------~------~--~
o 2 4 6
Time (hr)

Fig. 4. Mean plasma versus time concentrations of didanosine in ten patients following
a single 300-mg oral dose of a chewable tablet under fasting conditions or at various
times relative to a high-fat, high-calorie meal. (From KNUPP et al. 1993)

(SHYU et al. 1991). The results obtained in these two studies have led to the
recommendation that didanosine be administered under fasting conditions.
A milder food effect was reported for the novel quinazoline anti-
hypertensive agent doxazosin. Following single I-mg doses with a standard-
ized light breakfast, mean emax in 12 hypertensive subjects was reduced 18%
and the AVe by 12% compared to fasting. The acute fall in blood pressure
following doxazosine doses was not significantly affected by food, but there
was a trend toward lower instances of complaints after this treatment. Apart
from that, the modest food effects observed in this study are unlikely to be
clinically significant. A more serious clinical effect with flecainide was possibly
attributed to a milk drug interaction (RUSSEL and MARTIN 1989). Toxicity in
the form of ventricular tachycardia occurred in a baby boy when dextrose was
substituted for milk foods during flecainide therapy. Serum flecainide levels,
based on single point determinations, were doubled from 990.ug/ml to 1824.ug/
60 P.G. WELLING

ml 24 h following dextrose substitution. The seriousness of the effect would

justify close flecainide monitoring in these types of patients who may be on
intermittent milk diets. This has obvious implications for other dairy-based
diets in this population.
In an attempt to shed some light on variable results achieved in previous
studies that have reported increased, unchanged, and decreased hydralazine
absorption with foods, a study was conducted to compare the effect of fasting
and isocaloric standard breakfast, bolus enteral nutrients, and infused enteral
nutrients (SEMPLE et al. 1991). The study was conducted in four healthy men
and four healthy women. Both the standard breakfast and bolus enteral nutri-
ent treatments gave rise to much lower blood levels of hydralazine than fasting
and infusion of nutrients, although there was considerable interindividual
variation. Typically the fasted treatment yielded Cmax and AUC(0-6) values of
87 nglml and 2641 ng'min/ml compared to 11 nglml and 999 ng·min/ml and
15ng/ml and 1189ng'min/ml for the standard breakfast and enteral bolus
treatments, respectively. While not addressing the underlying cause of the
conflicting results reported previously, this study does show that, for
hydralazine at least, the physical form of nutrients may not play an important
role in determining the extent and nature of drug-food interactions.
An interesting result was obtained when plasma levels of levodopa, 3-
D-methyldopa, and carbidopa were compared in patients with idiopathic
Parkinson's disease after taking Sinemet CR, a controlled release formulation
containing 200/50 levodopa and carbidopa (Roos et al. 1993). As might have
been expected from previous studies, coadministration with a high-protein
meal reduced plasma levels of both levodopa and carbidopa due, amongst
other things, to competition for active absorption mechanisms. However, con-
centrations of the levodopa metabolite 3-D-methyldopa increased in the pres-
ence of the high-protein meal. Increased metabolite levels may be related
to increased gastric residence time in the fed state, giving rise to greater
presystemic conversion of levodopa to the metabolite. While levodopa and
carbidopa levels were generally reduced by the high-protein meals, the plasma
levels yielded a flattened concentration-effect profile. If a "kick" from
levodopa is required, then administration on an empty stomach may be
appropriate.
In a study in 14 juvenile rheumatoid arthritis patients the mean systemic
availability of methotrexate was reduced by 15% and Cmax by 38% when
medication was given immediately after breakfast compared to fasting
(DUPUIS et al. 1992). Compared to i.v. administration, the mean absolute
bioavailability of methotrexate in these patients was 90% in the fasting state
compared to 77% after breakfast. Food has also been shown to reduce sys-
temic availability of the investigational LTD4 receptor antagonist MK-679
(YOGENDRAN et al. 1992). Administering the drug after a standard meal re-
duced both Cmax and AUC(0-12) by approximately 40%.
Reduced nitrendipine absorption due to food is reported in an efficacy
study in juvenile patients with severe hypertension (WELLS and SINAIKO 1991).
Drug-Food Interactions Affecting Drug Absorption 61

Nitrendipine failed to lower blood pressure in only 1 of 25 patients. Plasma

levels of nitrendipine in 13 patients were variable but opportunity was taken to
compare levels in fed and fasted patients. Cmax values were reduced from 71 to
13ng/ml and AVCss values from 158 to 46ng·h/ml while t max was increased from
1 to 2.4h by food. Although the food effect probably contributed to variability
in nitrendipine levels, this was not further discussed, nor was the basis used to
differentiate fed and fasted subjects.
As noted previously with ciprofioxacin (NEUVONEN et al. 1991), co-
administration with milk or yoghurt had a marked effect on systemic availabil-
ity of norfioxacin (KIVISTO et al. 1992). In what appears to be emerging as a
possible class effect, both the plasma Cmax and the AVC(0-24) of norfioxacin
were reduced by approximately 50% when given with 300ml milk or yoghurt
compared to 300ml water. Plasma profiles obtained in this study are shown in
Fig. 5. The results of this study provide further evidence that, while absorption
of oral fiuoroquinolones is not affected to a clinically significant extent by solid
foods (HOLMES et al. 1985; MONK and CAMPOLI-RICHARDS 1981; VANCE-BRYAN
et al. 1990), dairy products containing solubilized calcium may reduce circulat-
ing drug levels to an extent that may well be clinically significant depending on
the organism(s) being treated.
A further example of the different effects by different diets on absorption
is provided by paracetamol, whose absorption was incomplete and abnormally
slow in Thai vegetarian compared to nonvegetarian subjects (PRESCOTT et al.
1993). In another study, however, the absorption of tetracycline was similarly
depressed by a Mexican meal and also by a conventional Western meal,
although the Mexican meal contained double the amount of calcium (COOK et

c
·0
ro
)(
0 0.4
...0
'=
c
ro
E
<II
ftI
0.2
'6:

0
0 1 2 3 4 6 8 10
~24
Time (hr)

Fig. S. Mean plasma versus time concentrations (±SE) of norfloxacin in seven subjects
following a single 200-mg oral dose of norfloxacin with 300ml milk (e), yoghurt (.), or
water (0). (From KIVISTO et al. 1992)
62 P.G. WELLING

al. 1993). Clearly both diets in this study contained sufficient free calcium to
markedly reduce systemic availability of tetracycline.
Contradictory effects were observed on pravastatin pharmacokinetics and
pharmacodynamics when the drug was taken with meals (PAN et al. 1993). In
a study conducted in 24 hypercholesterolemic men, mean Cm• x dropped by
49% and AUC by 31 % in the presence of food compared to the fasting state.
However, reduction in mean total cholesterol and low-density lipoproteins
was identical whether pravastatin was taken with or before food. The reduc-
tion in pravastatin bioavailability in these hypercholesterolemic patients was
similar to that observed in a preliminary study in healthy volunteers. The lack
of effect on blood lipids may have been due to food increasing extraction of
pravastatin by the liver, which is also the primary site of cholesterol synthesis,
LDLIcholesterol clearance, and also pravastatin activity. Other studies within
this category have reported 60% reduction in the already poorly absorbed
renin inhibitor remikiren (WEBER et al. 1993), a modest reduction in the
bioavailability of the leukotriene receptor antagonist SK&F 106203 (NICHOLS
et al. 1992), and moderately reduced plasma levels of rufloxacin (SEGRE et al.
1992) in the presence of food. In the last case, however, although mean Cm• x
and AUC values were decreased by 16% and 33%, respectively, urinary
excretion of unchanged drug increased from 16% to 26% of the administered
dose. The increase of urinary excretion suggests that lower plasma levels of
rufloxacin may not have been due to reduced drug absorption.
The beta-blocking agent sotalol appears to behave more like atenolol
(BARNWELL et al. 1993) than metoprolol or propranolol in that its systemic
bioavailability is reduced by approximately 20% by administered food. Like
atenolol, sotalol undergoes little or no first-pass metabolism so that any
change in hepatic blood flow due to food is not likely to have a positive effect
on systemic availability.
Formulation was shown to make a negligible contribution to the nature of
drug-food interaction for the neoleptic agent sulpiride (SHINKUMA et al. 1990).
Ingestion of a film-coated tablet or a solution with a standard meal resulted in
30% reduction in systemic availability. Further studies with the solution dos-
age form showed that the food effect increased with increasing meal size,
inhibition increasing from 21 % with a light meal, to 30% with a medium meal,
to 40% with a heavy meal. The systemic availability of the cognition agent
tacrine was reduced by approximately 20%, and peak plasma levels by 40%
when administered up to 2h following the start of a standard breakfast (WELTY
et al. 1994). However, administration with meals reduced gastrointestinal side
effects, thus improving patient tolerance. In the case of verapamil, food has
been shown not only to reduce systemic availability from a sustained release
formulation, but also to alter electrographic effects following a single drug
dose. There was no evidence of exaggerated release or "dose dumping" in this
study.
A number of studies have examined the effect of food on the absorption
and pharmacokinetics of the HIV agent zidovudine (LOITERER et al. 1991;
Drug-Food Interactions Affecting Drug Absorption 63

UNADKAT et al. 1990; PUHNKE et al. 1993; UEDA et al. 1993). All of these
studies, conducted in laboratories in Japan, Germany, and the United States,
have reported significantly delayed and reduced absorption with remarkable
consistency across studies. Typically, Tmax is increased from O.S to l.S-2.Oh,
while Cmax and AUC are decreased by approximately SO% when zidovudine is
administered with a meal compared to fasting. Variability in plasma levels is
also increased by food. While no direct relationship has been established
between circulating levels and efficacy of zidovudine, doses found to be effec-
tive in AIDS patients generally achieve peak serum levels of ~0.27 ,ug/ml.
However, lower and more prolonged levels, as observed when zidovudine is
taken with meals, may maintain efficacy while reducing toxic side effects
(UNADKAT et al. 1990).
For the last compound in this category, 882C (1-(j3-D-arabinofuranosyl)-S-
(1-propynyl)uracil), which is under investigation for treatment of varicella
zoster virus infections, food caused a significant reduction in Cmax and a modest
reduction in AUC in healthy volunteers (PECK et al. 1993). The effects ob-
served in this single-dose study led to the recommendation that effective
concentrations of this agent can be maintained without food-related
restrictions.

E. Interactions Causing Delayed Drug Absorption

Compounds whose absorption is delayed by food are listed in Table 4.
The large number of compounds in this category reflects the frequency of
this phenomenon, due for the most part to delayed gastric emptying and
dissolution.
Delayed absorption may be considered to be unimportant for the majority
of drugs, particularly when absolute systemic availability is not altered. This
may be particularly true during repeated doses, during which time circulating
drug profiles in the nonfasted state would be simply offset in time relative to
those in the fasted state, and overall profiles would tend to be somewhat flatter
as a result of reduction in peak drug levels. This may be the case with many
drugs and formulations included in this category in this chapter. However, as
in all the other categories described in this chapter, the importance of an
interaction depends on the extent of change compared to the required circulat-
ing profile and the therapeutic index of the drug. Peak circulating levels of
some of the drugs included in Table 4 are reduced by as much as 60% with
concomitant food ingestion. Changes of this magnitude in circulating drug
levels are likely to be therapeutically important. Other cases of more mod-
erate delays in absorption are likely to be less important therapeutically.
Typically, observed delays in effects or circulating profiles of acetorphan
(BERGMANN et al. 1992) and albuterol (HUSSEY et al. 1991) due to food are
probably clinically unimportant whereas the SO% reduction in Cmax reported
for aniracetam by food is likely to have clinical consequences (RONCARI 1993;
Table 4. Drug-food interactions resulting in delayed drug absorption
Drug Dosage form Dose Dosage regimen Subjects Food details Fluid Time interval Sampling Result of food Reference
volume duration administration

Acetorphan l00mg SOb 9 healthy subjects Standard meal, 20 min after meal - T min for enkephalinase BERGMANN
800kcal activity 180 min after food et al. (1992)
compared to 80 min for
acetorphan alone
Albuterol Volmaxand 4mg SO 23 healthy male Standard 250ml water Immediately after 30h plasma emu reduced 19% for HUSSEyet a1.
Proventil subjects breakfastc meal Volmax and 21 % for (1991)
Repetabs Proventil. Tmax and tlag
Albuterol sulfate delayed
5-Aminosalicylic Tablet SOOmg MDd 12 healthy male Breakfast, 100 ml water Immediately after 48h plasma Systemic absorption delayed DE MEyand
acid subjects afternoon snack, meals and urine but not reduced by food MEINEKE (1992)
evening snack,c
standardized
Aniracetam Tablet 1200mg SD emax reduced 50%, TIDaJI. RONCARl (1993)
extended from 0.6 to 1.1 h. HONMA et a1.
AVC of metabolites (1986)
unaffected by food
Beta- Lanirapid 0.2mg SO 9 healthy male Standard breakfast' 150ml water 30 min before or 120h serum, Absorption rate significantly TSUTSUMI et aI.
methyldigoxin subjects 560 kcal. Other after meal 240 h urine reduced after meal. Mean (1992)
meals also Cmax reduced 46%, Ave
controlled unaffected. V rinary
excretion slightly decreased
Cefaclor Capsules 500mg SD 8 healthy male Two standard 30 min after meal 6 h plasma IillIX
C reduced, Tmax OGUMA et a1.
subjects breakfasts. cOne and urine prolonged. Effect increased (1991)
based on rice, the with larger meal size. Rice
other based on eggs meal affected absorption
and bread. Six more than bread meal
months later
repeated, rice meal
reduced, bread
meal increased
Cefdinir 10% fine 3 mg/kg SD 8 children 30 min before or Cmax reduced 43%. Tmax MOTOHIRO et aI.
granules after meal increased from 2 h to 4-5 (1992)
h. AVC reduced by 15%.
Urinary recovery marginally
reduced
Cefdinir 5% fine granules 3 mg/kg SD 20 schoolchildren - 30 min before Cmax reduced 30%, Tmax NAKAMURA and
(14 children) or increased from 2 to 3.67 h. IWAI (1992)
30 min after V rinary recovery unchanged
(6 children) meal
Cefdinir 5% fine granules 3 mg/kg SD 25 younger 30 min before (19 - CroaK reduced 52%, Tmax NAKAMURA and
children children) or 30 increased from 2.1 to IWA! (1992)b
min after (6 3.3 h. Urinary recovery
children) meal reduced by 48 %
Cefdinir 5% fine granules 3 mg/kg SD 15 infants 30 min before (7 No significant differences NAKAMURA and
infants) or 30 min between groups IWA! (1992)
after (8 infants)
meal
Cefprozil Capsule 100mg SD 15 healthy male Standard breakfast' 200 ml water 30 min after meal 12h plasma, Tmax increased from 1.5 to SHUKLA et a1.
subjects 24h urine 2 h. Cm~' AUC, urinary (1992)
excretion not affected
Cefprozil Capsule 250mg SD 12 healthy male Standard breakfast C 150 ml water 15 min after meal 8h plasma Slight reduction in emu' BARBHAIYA
subjects mIDI.
Mean T increased from et a1. (1990)
1.2 to 2h
Cefaclor Capsule 250mg SD 12 healthy male Standard breakfast' 150ml water 15 min after meal 5 h plasma CmllJr reduced 51 %, Tmax BARBHAIYA et
subjects increased from 0.6 to 1.3 h a1. (1990)
CL 275838 Tablet 50mg SD 10 healthy male Standard breakfastC 150ml water 30 min after meal 48h plasma Tm~ delayed 2 h. Cm~' CONFALONIERI
subjects AUC unaffected. et a1. (1992)
Dic10fenac Capsule 50mg SD 8 healthy male Standard breakfast 150ml water Together with 10 h serum e max reduced 61 %, Tmu TERHAAG et a1.
subjects meal increased from 0.8 to (1991)
2.4h. AUC unaffected
Diclofenac Hydrogel bead 150mg SD 12 healthy male Standard breakfastC 200ml water 30 min after meal 24h plasma Cmax reduced 38 %, Tmax. THAKKER et al.
capsule subjects increased from 1.85 to (1992)
6.25 h. AUC unaffected
Diltiazem Pellets in capsule 240mg SD 8 healthy subjects Standard breakfast, 100ml water After meal 24h plasma Tmax delayed from 7.4 to WILDING et a1.
4M,4F lunch, dinner: 9.7 h. C_ and AUC (1991a)
unaffected
Doxycycline Pellets in capsule 100mg SD 20 healthy male Standard breakfast 180 ml water With breakfast 48 h plasma Lag time increased from WILLIAMS et al.
subjects and urine 0.55 to 1.1 h, Tmuincreased (1990)
from 1.8 to 3.1 h. C=" AUC,
and urinary excretion not
significantly affected
Erythromycin Enteric-coated 400mg b.i.d. SDandMD 14 healthy Standard light and Immediately after 12 h plasma After SD, both meals JARVINEN et a1.
acistrate tablet for 4 days subjects 4M, lOF heavy breakfastC meal delayed absorption. In seven (1992)
subjects no absorption
occurred in 12 h. After
MD food effect diminished
Fadrozole Tablet 12mg SD 9 healthy Standard breakfastC 240 ml water Within 20 min of 36 h plasma Cmax reduced by 15%, CHOI et al.
Caucasian completing meal T max increased from 1.8 to (1993)
SUbjects, 7M, 2F 2.5 h. AUC not affected
Table 4. Continued
Drug Dosage form Dose Dosage regimen Subjects Food details Fluid Time interval Sampling Result of food Reference
volume duration administration

Famotidine Tablet 40mg SD 6 p.m. or 10 healthy Supper given at With meal Variable Absorption delayed MARGALITH
9 p.m. subjects, 4M, 6F 6 p.m. or 9 p.m. plasma somewhat when meal and et aI. (1991)
drug coadministered at
6 p.m. No delay when meal
and drug coadministered at
9 p.m.
Flurbiprofen Tablet 100mg SD 12 healthy male 30 ml apple juice' 180ml water After the apple 48 h serum Gastric emptying time DRESSMAN
subjects juice almost doubled by meal. et al. (1992)
Slower initial absorption
followed by more rapid
second absorption phase
Fluvastatin Solution or 10mg SD 16 healthy male Standard high-fat Immediately 12 h blood e max reduced 69% to 74% SMITH et al.
capsule subjects breakfast after meal and TmaX increased more (1993)
than fourfold from solution.
Cmax reduced 50%--60%
and TmaX increased three-
to fourfold from capsule.
Bioavailability decreased by
15%-25%
Fluvastatin Capsule 20mg SD 22 healthy male Carbonated With carbonated Cmax reduced 57%, Tmax SMITH et a1.
subjects beverage beverage increased from 0.71 to 1.11 (1993)
h. Bioavailability reduced
28%
Fluvastatin Capsule 20mg SD Healthy male Low-fat evening 4 h after meal 12 h plasma Cmax reduced 44%, Tmax SMITH et al.
subjects meal (6 p.m.) increased 57 %. (1993)
Bioavailability reduced 14%
Fusidate sodium Film-coated 500mg SD Standardized After meal 26 h serum Cmax reduced 26%, Tmax MACGowAN
tablet breakfast increased from 2.2 to 3.2 h. et al. (1989)
AUC marginally
reduced (16%)
H ydroxychloro- Tablet 155mg SD 9 rheumatoid Light breakfast After meal Absorption delayed due to McLACHLAN
quine arthritis patients food to similar extent to and CUTLER
healthy subjects (1991)
Isosorbide-5- Controlled 60mg SD 18 subjects High-fat breakfast Cmax reduced 5%, Tmax KOSOGLOU
mononitrate release tablet increased from 3.1 to 6.5 et al. (1993)
h. AVe increased 9%
Lomefloxacin Capsule 400mg SD 12 healthy High-fat or high'- 200ml water 5 min after meal 48h plasma, Tmax delayed approximately HOOPER et a1.
subjects, 8M, 4F carbohydrate 72 h urine I h, em" and AUC (1990)
breakfast unchanged
Loracarbef Capsule 400mg SD 12 healthy male Standard breakfaste 200ml water 5 min after meal 24 h serum Cmax reduced 29%, Tmax ROLLER et a1.
subjects fasting, 100 and urine increased from 68 to 141 min, (1992)
ml water fed AUC unchanged
Loracarbef Capsule 200mg SD 12 subjects 12 h plasma Cmax reduced and Tmax DESANTE and
delayed. AUC unchanged ZECKEL (1992)

Methotrexate Tablet 7.Smg SD 12 healthy male Standard high-fat 10 min after meal 24h plasma Cmax reduced by 16%, KOZLOSKI et al.
subjects breakfast e Tmax delayed from 1.0 to (1992b)
1.6 h. AUC increased 4.6%
Methotrexate Tablet 1Smg SD 10 patients, SM, Standard French After meal 24h plasma Cmax reduced 31 %, Tmu OGUEyet a!.
SF breakfast' increased from 1.3 to 2.0 h. (1992)
Bioavailability unaffected
Monofluoro- Chewable tablet 10mg SD 8 healthy male Standard meale 200ml Immediately after 48 h plasma Lag time delayed from 4 WARNEKE and
phosphate fluoride subjects fluoride-free meal and urine to 11 min, Cmax reduced by SETNIKAR
with 300mg water 67%. Tmax increased from (1993)
calcium 24 min to 2.4 h. AUe and
urinary excretion not
affected
Moricizine Tablet 2S0mg SD 24 healthy male Standard breakfaste 180ml 30 min after meal 24h plasma Cmax reduced 24%, Tmax PIENIASZEK
subjects orange juice delayed from 0.9 to 1.2 h. et aJ. (1991)
AUC not affected
Nicorandil Tablet 10mg SD 10 healthy male Standard breakfast' - 10 min after meal - Cm~ reduced 13%, Tm~ FRYDMAN
subject increased from 0.38 to (1992)
0.73 h, AUC increased 24%.
Absorption rate consistently
decreased by 90%
Nifedipine 10mg SD 10 healthy subjects High-fat meal 100ml water After meal 24h plasma Cmax reduced by 47%, Tmu RAPIN (1992)
increased from 0.9 to 1.44 h,
AUC unaffected. Under
same study conditions of
nicardipine and nitrendipine
unaffected
Ofloxacin Tablet 200mg SD 7 healthy male 300ml milk 300 ml water Immediately 24 h plasma Cmax reduced 18%, Tmax NEUVONEN and
subjects 300 ml yoghurt' before milk and urine delayed 1 h by yoghurt. KlVlSrti (1992)
or yoghurt AU C unaffected. Milk had
no effect on any parameters
Ofioxacin Capsule 300mg SD 21 healthy male Standard breakfast' - 30 min after food 24 h plasma Cmax reduced 17%, Tmax DUDLEY et a1.
subjects or milk or milk delayed 1 h by food, AUC (1991)
unaffected. Milk had no
effect on any parameters
Table 4. Continued
Drug Dosage form Dose Dosage regimen Subjects Food details Fluid Time interval Sampling Result of food Reference
volume duration administration

Paracetamol Tablet and 500mg SD 12 healthy male 200mlliquid 200ml With liquid diet 12h plasma The two liquid diets did WALTER-SACK
solution subjects formula enriched not differentially affect et al. (1989)
with or depleted of absorption from solution.
dietary fiber' Fiber-depleted diet delayed
absorption from tablets
relative to fiber -enriched diet
Paracetamol Tablet l000mg SD 12 healthy Tswana Cereal, maize, or 200ml water 30 min after meal 12 h serum Absorption delayed to WESSELS et a1.
female volunteers bacon and egg increasing extent by maize, (1992)
breakfastC bacon and egg. and cereal
breakfasts
Penciclovir 250 or SD 12 healthy male 24h plasma Cmax reduced 47% and FOWLES et a1.
(famcic1ovir 500mg subjects 45%, Tm" delayed by 1.7 (1990,1991)
prodrug) hand 1.8 h from 250- and
500-mg doses, respectively,
AUe unaffected
Penciclovir 500mg SD 18 healthy male 2 h before or 24h plasma emax reduced 18%, Tmax FOWLES et aI.
(famcic10vir subjects after meal and urine delayed by 1.1 h (1990, 1991)
prodrug)
Rifabutin Capsule 150mg SD 12 healthy male Standard high-fat 15 min after meal 168hpiasma, Cmaxreduced 17%, Tmax NARANG et a1.
subjects breakfast' 48 h urine delayed 2.4 h, Aue (1992)
unaffected. Urinary
excretion reduced 20%
Salsalate Tablet 1500mg SD 17 healthy male Standard high-fat 6 ounces Immediately after 48h plasma Salsalate mean Cmax reduced HARRISON et al.
subjects breakfastC water meal 17%, Tmu delayed 1 h, AUe (1992)
unaffected. Salicylate
parameters unaffected
Terazocin 2mg SD 12 subjects Standard breakfast' 100 ml water - 30h plasma Mean Cmax reduced 23 %. McNEAL et al.
Tmu delayed 0.9 h, AUe (1991)
unaffected
Terfenadine Tablet 120mg SD 24 healthy male Standard breakfast' 200ml water Immediately after 48 h plasma Mean metabolite 1 Cmax ELLER et at.
subjects meal increased 13%, TmaJ[delayed (1992)
by 0.9 h. AUe unaffected
Theophylline Theo-Dur tablets 400mg SD 10 healthy male High-fat dinner' 6 ounces Immediately after 36h serum Mean Cmax increased 33 %, KANN et al.
subjects water meal, 8 p.m. T ron unchanged, AU e (1989)
increased 6%. Absorption
rate delayed during initial
8 h postdosing
Theophylline Multiparticulate 300mg SD 12 healthy male Standard breakfastC 150ml water Immediately after 36h serum Delayed drug absorption YUEN et a1.
CR pellet subjects meal associated with delayed (1993)
stomach emptying
Tiagabine Tablet 8mg SD 18 healthy male Cmwt reduced 44%, Tmu MENGEL et a1.
subjects increased from 0.9 to 2.6 h. (1991)
AUC unaffected
Topiramate Tablet 100mg, SD 18 healthy male Fatty breakfast 168 h plasma Cm.. reduced 11 % (100 mg) DOOSE et al.
400mg subjects and 13% (400mg), Tmu (1992a)
increased by 1.8 h (100 mg)
and 2.1 h (400 mg), AUC
(0-=) unaffected
Trazodone Capsule 100mg SD 8 healthy subjects, Standard breakfast After meal 26h serum Cmax reduced 22%, T JDBJ. NILSEN and
4M,4F and urine increased from 1.3 to 2.0 h. DALE (1992)
Amount of drug absorbed
unchanged
Valproic acid Tablet 800mg SD 16 healthy male Light meal, heavy" 180 ml water After meal 45 or 48h emu reduced 14% and AVe OHDO et al.
subjects meal plasma 9% by heavy evening meal, (1992)
but not affected by light
evening meal relative to
light breakfast
Vigabatrin Tablet 1000mg SD 24 healthy male 36 h plasma, Cmax reduced 33%, Tmax HOKE et al.
subjects 48 h urine increased from 1.0 to 2.14 h. (1991)
Extent of absorption
unaffected
Zalospirone 20mg SD 24 healthy male High-fat, low-fat After meal 24h plasma Cmu reduced 13% by low-fat KORTH-
subjects breakfast and 31 % by high-fat meals. BRADLEY et a1.
Tmax increased 0,5 hand (1992)
1.2 h. AU C unaffected
Zidovudine 100mg SD 18 asymptomatic High-fat breakfast 30minor3h 10h plasma Cmax reduced 57% (30 min) SHELTON et a1.
HIV -infected after meal and 47% (3 h). Tm .. (1993)
subjects increased by 1 h (30 min)
and 0.6 h (3 h). AUC
reduced 3% (30 min) and
15% (3 h)
Zidovudine Capsule 200mg SD 11 symptomatic Liquid protein meal 150 ml water Immediately after 8h serum emu reduced 32%, TrnaJ. SARAI et a1.
HIV-infected in 220 ml orange meal and urine increased from 49 to 74 min. (1992)
men juicec AVe and renal clearance
unaffected

a Data not available.

b Single dose.
C Meal described.
d Multiple doses.
 70 P.G. WELLING

HONMA et al. 1986). Delayed absorption of 5-aminosalicylic acid (5-ASA) by

food was confirmed in a study in which repeated 500-mg oral doses of 5-ASA
were administered to healthy male subjects either 1 h before or immediately
after meals (DE MEY and MEINEKE 1992). A surprising observation in this
study was a marked peak in plasma drug concentrations in the early morning
following repeated doses the day before, regardless of dosing time relative to
food. Mean values are shown in Fig. 6. This dramatic food-independent effect
is postulated to be the result of tablets, or tablet fragments, remaining in the
stomach until early morning, at which time they are released quite suddenly as
a result of interdigestive phase 3 contractions occurring at that time of day.
Lack of adequate interdigestive housekeeping waves during the previous day
may have been due to snacks and meals shortening the interdigestive phase.
Possible clinical significance is claimed for the delay in beta-methyldigoxin
absorption in the presence of food (TSUTSUMI et al. 1992). Mean peak serum
beta-methyldigoxin levels were delayed and reduced from 2.lDng/ml and
1.79ng/ml when drug was taken 30min before a standard breakfast and fast-
ing, respectively, to O.96ng/ml when drug was taken 30min after food. The
results obtained with beta-methyldigoxin, which are similar to those reported
in some studies for digoxin (SANCHEZ et al. 1973; JOHNSON et al. 1978), are
summarized in Fig. 7.
A number of studies have examined and compared the effects of ingested
food on circulating levels of oral cephalosporins. Absorption of cefaclor was

40
aJ
~ 36 aJ '"0 aJ
S '"0
"C
"C
C '"0
"C
~ 32
.
0

.
OIl OIl
C 0
,;:;. c c
'2 '2
aJ
:2 28 0 .:t: aJ
\J :; ...: >
'"
L.L.I

.~ 24
>..
~ 20
'"'c0"
·s 16

.;.,'" 12
'"
E 8
''""
"6:
4

0
-1 3 7 11 15 19 23 27 31 35 39 43 47
07.00 14.00 21.00 07.00

Time (hr) and real dock time (hr)

Fig. 6. Median plasma versus time concentrations of S-aminosalicylic acid in 12 sub-

jects following three oral administrations of 3 x SOO-mg S-aminosalicylic acid tablets
either 1 h before (0) or at the end of (e) meals. (From DE MEY and MEINEKE 1992)
 Drug-Food Interactions Affecting Drug Absorption 71

3T---------------------------------------------------,

~
5
c

0
..
·0
IU

;;
l:
c..
·0
IU
E
'"
IU
E:
0

-1+---~--~--~--_r--_r--~------_,--_,--_,r__,r_--~
o 2 4 6 8 10 12
Time (hr)

Fig. 7. Mean serum versus time concentrations (±SD) of beta-methyldigoxin in nine

subjects following a single 0.2-mg oral dose of beta-methyldigoxin fasting (0), 30min
before (D) and 30 min after (e) a standard breakfast. (From TSUTSUMI et al. 1992)

delayed but not reduced following rice and bread meals, with rice having a
greater effect than bread (OGUMA et al. 1991). For both meal types, the effect
was greater for a large meal than for a small meal. Similar delays in cefaclor
absorption were observed when drug was taken immediately after a standard
breakfast, resulting in a 51 % reduction in mean peak plasma levels
(BARBHAIYA et al. 1990). Under the same study conditions peak levels of
cefprozil were reduced by only 14% and plasma profiles were similar after
fasting and nonfasting treatments. The different effects on plasma profiles of
cefaclor and cefprozil are shown in Fig. 8. It is claimed that because cefaclor is
absorbed more rapidly than cefprozil under fasting conditions, slight perturba-
tions in stomach emptying and GI motility are more likely to affect the
absorption rate of cefaclor than of cefprozil. Other studies have shown
cefadroxyl to be unaffected by food, similarly to cefprozil, while absorption of
cephalexin is delayed similarly to cefac10r (LODE et al. 1979). In further studies
designed to examine the influence of gastric emptying on cefprozil absorp-
tion, food delayed the cefprozil tmax in plasma from 1.5 to 2h but did not
affect peak cefprozil plasma levels or absorption efficiency. Pretreatment
with metoclopramide reduced the tmax and mean residence time while
propantheline increased the tmax and significantly increased mean residence
 72 P.G. WELLING

8 8

7 7

-
E
6
-
E
6

~5 "Ob
.
5
2-
·N
e
0
4 1:i 4
~
~ <II
'!IS"' '"'
!IS
E 3 E 3
'"
!IS '"
!IS
ii: ii:
2 2

o~----~------~----~----~ O~----~------~----~-------i
o 2 4 6 8 o 2 4 6 8

Time (hr)
Fig. 8. Mean plasma versus time concentrations of cefprozil and cefaclor in 12 subjects
following single 250-mg oral doses of cefprozil or cefaclor under fed (e) and fasting
(0) conditions. (From BARBHAIYA et al. 1990)

time. No other pharmacokinetic parameters were affected. Thus, the rate of

cefprozil absorption is modestly influenced by gastric emptying rate but the
extent of absorption is not affected. Peak plasma levels of cefdinir were
reduced 30%-S2% by food in children of various ages who received drug as a
S% or 10% fine-granule suspension (MOTOHIRO et al. 1992; NAKAMURA and
IWAI 1992b). Interestingly no reduction in Cmax occurred; in fact the mean Cmax
value increased from 0.61.ug/ml in the fasting state to 0.79 .uglml in the fed state
in infants who received a S% fine-granule 3-mg/kg pediatric dose.
Co administered food has been shown to have a marked effect on circulat-
ing levels of the nonsteroidal anti-inflammatory agent diclofenac. Absorption
from a 1S0-mg hydrogel bead capsule in healthy male subjects was delayed,
yielding a mean Cmax value of 312nglrnl at 6.2Sh compared to S02nglmi at 2h
in the fasted state (THAKKER et al. 1992). Systemic availability was not signifi-
cantly affected by food but was SO% lower from the capsule dosage than from
a solution. In an attempt to elucidate the mechanism of delayed diclofenac
absorption, TERHAAG et al. (1991) compared diclofenac absorption in the
presence and absence of food using paracetamol as a marker of gastric empty-
ing rate. Under the conditions of this study, food had no effect on the rate or
extent of paracetamol absorption, leading to the conclusion that delayed ab-
sorption of diclofenac due to food may not be due to delayed stomach empty-
ing alone, but rather some interaction between the drug and food components,
Drug-Food Interactions Affecting Drug Absorption 73

possibly adsorption, resulting in delayed stomach emptying of drug together

with food. The results observed in these two studies using diclofenac capsules
are in contrast to those using diclofenac enteric coated tablets in which peak
circulating levels were delayed but not reduced by food (TERHAAG et al. 1990;
WILLIS et al. 1981). The capricious relationship between stomach emptying
and drug absorption rate was demonstrated with the calcium channel blocking
agent diltiazem (WILDING et al. 1991a). Using gamma scintigraphy in healthy
subjects, gastric emptying Tso% values of sustained release diltiazem pellets
were significantly increased from 91 min fasting to 232 min with food, but the
time of peak plasma levels was increased only moderately from a mean value
of 7.4h to 9.7h. Other pharmacokinetic parameters were unaffected.
Results obtained in a food-effect study with erythromycin acistrate (2'-
acetyl erythromycin stearate) appear to be consistent with earlier reports from
other studies on erythromycin and its derivatives (JARVINEN et al. 1992). Food
has been shown to generally reduce the systemic availability of erythromycin
and erythromycin stearate (WELLING and TSE 1992) while absorption of eryth-
romycin esters is either unchanged or increased (WELLING et al. 1979). The
rate of erythromycin acistrate absorption from single doses was delayed to a
small extent following a light meal and to a greater extent following a heavy
meal. While the effect was attenuated to some extent after repeated doses,
food significantly delayed absorption in some individuals after single and
repeated doses, resulting in undetectable levels up to 12 or 24h postdosing.
Data supporting the influence of stomach emptying time on drug absorption
rate are provided by a study in which flurbiprofen tablets were administered
together with equal volumes of water or calorie-dense apple juice (DRESSMAN
et al. 1992). By use of a Heidelberg capsule, apple juice significantly increased
mean gastric emptying time from 57 min in the fasted state to 102 min in the fed
state. Loo-Reigelman analysis of the resulting serum flurbiprofen data showed
a tendency toward lower initial absorption rate followed by a more rapid
second phase when subjects received apple juice. Double peaks were observed
in plasma profiles from both treatments, but these were less than those ob-
served following a solution flurbiprofen dose. Double peaks are speculated to
be related to the existence of two distinct absorption sites separated by a
region of relatively lower absorption (PLUSQUELLEC et al. 1987) or the result of
biphasic stomach emptying (CLEMENTS et al. 1978). An extensive range of
studies conducted in healthy male subjects showed that food can markedly
affect circulating levels of the HMG-CoA reductase inhibitor fluvastatin
(SMITH et al. 1993). Peak plasma levels were reduced by 69%-74% and Tmax
was increased fourfold by a standard low-fat breakfast following a solution
dose of fluvastatin, while peak levels were reduced 50%-60% and tmax values
increased up to fourfold from a capsule dose. A similar, but somewhat attenu-
ated effect was observed following a low-fat evening meal, with peak levels
being reduced by 44% and the time of peak increasing by about 50%. In a
third study, administering fluvastatin together with a carbonated beverage
reduced peak plasma levels by 57% and tmax was increased from 0.8 to 1.1 h. In
74 P.G. WELLING

these studies systemic availability was reduced by 15%-20%. Comparison of

efficacy data following the evening dose study showed no significant difference
in the reduction in total cholesterol and LDL-cholesterol from baseline be-
tween fed and fasted subjects. Moderate food-effects resulting in small delays
and/or reductions in peak circulating drug levels have been reported for
fusidate sodium (MACGOWAN et al. 1989), hydroxychloroquine (McLACHLAN
and CUTLER 1991), isosorbide-5-mononitrate (KOSOGLOU et al. 1993), and
lomefloxacin (HOOPER et al. 1990). A more impressive food effect is reported
for the oral beta-Iactam antibiotic loracarbef (DESANTE and ZECKEL 1992).
While systemic availability was similar in subjects who received 200-mg doses
in fasting and nonfasting states, peak plasma levels were reduced over 50%
and the time of peak was doubled by food. In the same study, the area under
the loracarbef serum curve was approximately doubled by co administered
probenecid due in part to an increase in the elimination t1/2 from 1 to 1.5 h. A
less pronounced effect was obtained in a study in healthy male volunteers in
which dosing after a standard meal resulted in a 29% reduction in Cmax and a
doubling of tmax compared to the fasting state. It is claimed that, despite the
lowering of serum levels of loracarbef by food, levels were nonetheless well
above MIC values for susceptible bacteria. Interestingly, administration of
food significantly increased urinary excretion of loracarbef so that overall
absorption efficiency may have actually been increased.
Two studies examined the effect of food on absorption rate and
bioavailability of methotrexate. The first of these examined the effect of a
standard high-fat breakfast on absorption of methotrexate from 7.5-mg tablet
doses of the sodium salt in healthy male volunteers. Postprandial administra-
tion under these conditions resulted in a modest 16% reduction in peak
plasma levels and a 30-min delay in t max • The conclusion was drawn that food
has no significant effect on the rate and extent of methotrexate absorption. A
somewhat greater effect was observed when methotrexate was given fasting or
after a standard French breakfast in patients with rheumatoid arthritis (OGUEY
et al. 1992). In this case, Cmax was reduced by 31 % and tmax increased by 40 min.
Absolute systemic availability varied from 28% to 94% relative to intravenous
methotrexate and was not influenced by food. Despite the somewhat greater
reduction in peak levels in this study, it was again concluded that oral
methotrexate can be given without regard to food intake. A far more dramatic
effect was obtained when fluoride in the form of monofluorophosphate was
administered after a standard meal compared to fasting (WARNEKE and
SETNIKAR 1993). Food caused a 67% reduction in peak plasma fluoride levels,
369 to 122ngiml, and increased tmax from 1h to over 2-5h. AVC and urinary
excretion values were essentially unchanged. Monofluorophosphate is known
to pass essentially unchanged through the stomach and to be rapidly absorbed
in the upper small intestine. It is hydrolyzed to fluoride in the intestinal
mucosa, portal blood, and liver (SETNIKAR and ARIGONI 1988). Thus, slow
absorption in this case is attributed to delayed stomach emptying. Plasma
fluoride profiles under fed and fasting conditions are shown in Fig. 9.
Drug-Food Interactions Affecting Drug Absorption 75

Time (hr)
Fig. 9. Mean plasma versus time concentrations (±SD) of fluoride in eight subjects
following a single lO-mg oral dose of fluoride as sodium monofluorophosphate fasting
(0) or after a standard meal (e). (From WARNEKE and SETNIKAR 1993)

A high-fat meal markedly reduced the absorption rate of the calcium

channel blocking agent nifedipine (RAPIN 1992). Peak plasma levels were
reduced by 47% and Tmax increased from 0.9 to l.4h after a high-fat meal
compared to fasting. The related compounds nicardipine and nitrendipine, on
the other hand, were unaffected by food. A much smaller food effect was
obtained in studies examining the effect of food, milk, and yoghurt on
ofloxacin absorption (DUDLEY et al. 1991; NEUVONEN and KIVISTO 1992). Milk
had no effect, yoghurt caused an 18% reduction and a moderate delay in peak
plasma concentrations, while a standard meal gave rise to 20% reduction in
peak plasma levels, again with a moderate, approximately 1 h, delay. Ofloxacin
is thus consistent with other fluoroquinolone antibiotics in that it is only
modestly affected by food, at least under the conditions studied, and can
generally be administered without regard to meal intake. Meal types have
been shown to have different and probably not clinically significant effects on
the rate of paracetamol absorption among South African Tswana ethnic
groups (WESSELS et al. 1992). Absorption was delayed to an increasing extent
by maize, bacon and egg, and cereal breakfasts. The bacon and egg meal
reduced the paracetamol emax from 16.9 to 12.6.ug/ml and increased Tmax from
2.9 to 9.5h following a single lOOO-mg paracetamol dose. The greatest delay by
the high-fat meal is consistent with the expected impact of this meal on
stomach emptying rate.
76 P.G. WELLING

Food has been shown to decrease the absorption rate of the salicylic acid
derivative salsalate (salicylsalicylic acid) (HARRISON et al. 1992). However, the
effect is modest, and there was essentially no effect on circulating levels of the
major metabolite salicylic acid. A similar modest effect has been reported for
terazosin, whose absorption was moderately delayed by food, giving rise to a
23% reduction in Cm• x and an 0.9-h increase in Tmax (McNEIL et al. 1991).
Consistent with this modest effect, the maximal fall in standing blood pressure
after food was similar to that in the fasting state. Similar minor effects were
observed with terfenadine, with peak levels of the active metabolite decreas-
ing only 13%, with an 0.9-h delay following a standard high-fat breakfast
compared to the fasted state in healthy volunteers (ELLER et al. 1992).
Two studies reporting delayed absorption of theophylline due to food
have contributed further to the wide spectrum of food effects that have been
reported for this compound and its many formulations. Absorption of theo-
phylline from Theo-Dur tablets was delayed to a small extent, but mean peak
levels increased from 4.7 to 6.3mg/ml from a 400-mg postprandial evening
dose compared to the fasting state. While these results are consistent with
those reported previously regarding delayed absorption, the marked increase
in Cmax ' together with observed intersubject variability in serum profiles ob-
tained following evening doses, may have implications for asthmatic patients
with nocturnal asthma who require consistent medication at night. In a second
study, the onset and rate of gastric emptying were both delayed by a standard
meal and the rate of theophylline absorption from a multiparticular controlled
release formulation was also delayed, but to a lesser extent (YUEN et al. 1993).
Calculations of the relative amounts of drug absorbed from regions of the GI
tract showed that, in fasted individuals, 9% was absorbed from the stomach
and 54 % from the small intestine. In fed individuals 14 % was absorbed
from the stomach and 47% from the intestine. Following both treatments
37%-39% of the absorbed dose was absorbed from the colon, showing this to
be an important absorption site, at least for sustained-release products of
theophylline.
Delayed absorption due to food has been reported for the gamma-
aminobutyric acid uptake inhibitor tiagabine (MENGEL et al. 1991). In a study
in healthy male volunteers, peak plasma levels were reduced by 44 % and Tmax
was increased almost threefold under nonfasting conditions compared to fast-
ing conditions. AUC(0-48) values were similar for the two treatments. Similar
observations were made for the structurally novel anticonvulsant topiramate,
although peak plasma levels were reduced only 11 % and 13 % from single 100-
mg or 400-mg doses and Tm • x was increased by only 1 h (DOOSE et al. 1992a).
Absorption efficacy was unaffected by food. Similar modest delays in absorp-
tion have been reported for the serotonin uptake inhibitor trazodone (NILSEN
and DALE 1992). A more complex study on valproate pharmacokinetics, on the
other hand, shed some light on the influence of food on apparent "circadian
rhythm" in absorption of this and possibly other compounds (ORDO et al.
1992). In a study conducted in healthy male subjects, valproate was adminis-
Drug-Food Interactions Affecting Drug Absorption 77

tered following conventional light breakfast and heavy evening meals and also
following identical morning and evening meals. There was no fasting treat-
ment in this study. Following the light breakfast-heavy evening meal doses,
absorption of valproate was delayed following the evening dose relative to the
morning dose, Cmax being reduced by 14% and AUe by 9%. When valproate
was administered following identical morning and evening meals, there were
no differences in the resulting plasma valproate profiles. The different results
in the two treatments were attributed to differences in meal sizes and
the results of the study led to speculation as to what extent different food
effects may have contributed to many "circadian rhythm" effects previously
reported for other drugs. Plasma levels obtained in this study are shown in
Fig. 10.
Different meals also influenced the extent of drug-food interaction with
the serotonin agonist/antagonist zalospirone. In a study in 24 healthy men,
administration of a single 20-mg zalospirone dose after a meal containing 43%
fat resulted in a 31 % reduction in peak circulating drug levels compared to the
fasting state. Administration after a meal containing 19% fat, on the other
hand, resulted in only a 13 % reduction. Peak drug levels were delayed twice as
long after the high-fat meal as after the low-fat meal. AUe values and
incidences of adverse events were similar for all dosage treatments.
Whereas previously cited reports have claimed delayed and reduced
absorption of the AIDS drug zidovudine (see Table 3), other studies have
reported only delayed absorption, with no influence by food on overall
absorption efficacy (SAHAI et al. 1992; SHELTON et al. 1993). In a study in 18
asymptomatic HIV-infected subjects, Cmax was reduced 57% and 47% when
zidovudine was administered 30min and 3h after a high-fat breakfast, with
moderate increases in Tmax , while overall absorption efficacy was unaffected
(SHELTON et al. 1993). In a second study (SAHAI et al. 1992), serum zidovudine
Cmax was reduced 32% and Tmax increased from 49 to 74min when drug
was administered after a liquid protein meal relative to the fasted state in
11 symptomatic HIV-infected men. Again, absolute absorption values
were unaffected. The different results obtained in these studies compared
to those described in Table 3 present another example of the variable
nature of drug-food interactions and the hazards of basing any conclusions
on only one study from one laboratory, using a single set of study
conditions.

F. Interactions Causing Increased Drug Absorption

Unlike the drugs and dosage forms described so far in this chapter, the rate
and extent of systemic availability of the drugs and dosage forms described in
this section are increased in the presence of food relative to the fasting state.
Publications cited in Table 5 represent a substantial portion of the total num-
ber of reports on drug-food interactions during the review period, and reflect
 -...J
00

100 100

E 90 90
~ A B
2- 80 80
"t:I
'0 70 70
..s
1.1
'0 60 60
.
Q.
-;;; 50 50
..
..s 40
E 40
'"..s 30 30
!
E:
20 20
i !!
10 " 10

0 I I
II
I I
"I
o I I I I I I 1/ T
0 2 3 4 12 24 33 36 o 2 3 4 6 12 24
Time (hr) Time (hr)
Fig.l0A,B. Mean plasma versus time concentrations (±SD) of valproic acid in 16 subjects following single 800-mg oral doses
of valproic acid following a light breakfast (e) or heavy dinner (_) (A) or following a light breakfast (e) or light dinner (_)
(B). (From ORDO et al. 1992)
""d
o
~
tn
t'"
t:
z
Cl
Table 5. Drug-food interactions resulting in increased drug absorption
Drug Dosage form Dose Dosage Subjects Food details Fluid volume Time interval Sampling Result of food administration Reference
regimen duration

Alprazolam Sustained 3mg 29 healthy Standard high-fat _b 1 h before, 48h blood Cmax increased 10%-17% WRIGHT (1992)
SD'
release tablet subjects meal immediately after food. AUC marginally
following, 1 h increased (::;;6%). Psychomotor
after, and 2 h performance decrement
after meal consistent with changes
in blood levies
Amiodarone Solution 300mg SD 10 healthy male Nutrient RealmentylC 600 ml water Drug 24h plasma, Amiodarone absorption PFEIFFER et al.
subjects solution at two administered in 2 h jejunal correlated with lipid (1990)
dilutions infused 120 nutrient solution aspirations absorption. Plasma levels
min at ligament of variable
Treitz
Amocarzine Film-coated 3 mg/kg MDd 20 male Copious breakfast 100 ml water 1 h after meal 10h plasma AUe increased 20%, TrnlU LECAILLON
breakable tablet Guatemalan and urine delayed 2 h, Cmu unaffected. et al. (1991)
patients infected T m" of metabolite CGP 13231
with Onchocerca delayed 2 h
volvulus
Amocarzine Film-coated 1200 mg SD 11 male patients Standard breakfastC 100ml water After meal 48 h plasma, Cmax and AUe of parent LECAILLON
tablet infected with 480rl44h drug increased threefold. et al. (1990)
Onchocerca urine Urinary excretion increased
volvulus from 0.11 % to 0.13% of dose.
C max , AVe and urinary
excretion of metabolite eGP
13231 increased threefold
Astemizole and Capsule 10 mg astemizole, SD 28 healthy Standard meal With meal 96h plasma Cmax for astemizole JALLAD et a1.
Pseudoephedrine 140mg subjects and metabolite (1991)
pseudoephedrine desmethylastemizole
delayed while AUC increased.
Pseudoephedrine unaffected
Atovaquone Tablet 500mg SD 18 healthy male High-fat breakfast' 45 min after 528h plasma C max increased fivefold, AVe ROLAN et al.
subjects meal increased threefold (1994)
Atovaquone Tablet 500mg SD 12 healthy male Toast, toast with 45 min after 336 h plasma C max and AUe both increased ROLAN et a1.
subjects low-fat or high-fat meal with increasing fat meal. (1994)
butter Following fat toast with 56 g
butter. emu and AVC
increased 5.6- and 4-fold,
respectively
Table 5. Continued
Orug Dosage form Dose Dosage Subjects Food details Fluid volume Time interval Sampling Result of food administration Reference
regimen duration

Bay·X·1005 Tablet 500mg SO 4 healthy subjects American breakfast emax and AVe increased BECKERMANN
2·fold and l.4·fold, et al. (1993)
respectively. Tmu reduced
from 4.4 to 2.2 h
Brofaromine Tablet 75mg SO B healthy male Standard breakfast 100ml Immediately 4Bh plasma emax and AVe increased DEGEN et al.
subjects mineral after food 1.2-fold. Tmax reduced from (1993)
water 3 to 2h
Buflomedil Tablet 600mg SD 111In_ 7 healthy male Standard light and 100ml water After meal 25 h serum, Cm8~ and AU C increased WILSON et a1.
labelled subjects heavy breakfasts' scintigraphy approximately 1.2·fold after (1991)
heavy meal relative to light
meal
Cefetamet Tablet 1000mg SO 6 healthy male emu and AVe increased BLOUIN and
pivoxil subjects 1.3·fold. Tmn delayed from STOECKEL
3.0 to 4.Bh (1993)
Cefetamet Tablet 1000 mg SO 16 healthy male Standard breakfast' Various With meal 24h plasma Cmu increased 1.2·fold, AUC TAM et aJ.
privoxi subjects and urine increased 1.1-fold when (1990)
taken 1 h after food compared
to 1 h before
Cefuroxime Tablet 250mg SO B or 12 healthy Cholecystokinin or Hyoscine butylbromide MAcKAyet aJ.
fasting male hyoscine mediated delayed gastric (1991)
subjects butylbromide, no emptying had no effect.
food Cholecystokinin-mediated
increased bile release caused
1.2-fold increases in
C_andAUC
CGP 43371 Capsule BOOmg SO 12 healthy male Standard breakfast C After meal 96h plasma Cmax and AU C increased SUN et al.
subjects 11· and 14·fold, respectively. (1994)
Tmax unaffected
Clarithromycin Suspension 7.5 mglkg SO 24 infants and Milk for patients 125 mg/5 ml 6h plasma Cmax for parent drug and GAN et aJ.
children with <2 years old, milk metabolite increased 1.3- and (1992)
pharyngitis, otitis and hash brown 1.1·fold. AUC increased
media, or skin potatoes for patients 1.4· and 1.1 ·fold
infections ;,2 years old
Clarithromycin Tablet 500mg SD 27 healthy male Standard breakfast 240ml water 30 min after 24h plasma Cmax for parent and metabolite CHU et a!.
subjects start of meal increased 1.5- and 1.1-fold, (1992)
AVC increased 1.25-
and 1.1-fold
Cyclosporine Chocolate 10 mg/kg SD 8 healthy Low-fat and high-fat Midway 24 h blood Based on blood data, GUPTA et al.
emulsion subjects, 4M, 4F breakfast' through high-fat and plasma bioavailability was 23% and (1990)
breakfast 42% after low-fat and high-fat
meals. Based on plasma data
bioavailability was 21 % and
79% after low-fat and high-fat
meals. Distribution volume
and clearance increased with
high-fat meals
Danazol Capsule and 100mg SO 11 healthy Standard breakfast 15 min after 36 h plasma Emulsion increased CHARMAN
lipid emulsion female subjects meal bioavailability fourfold et a!. (1993)
relative to capsule, but
was unaffected by food.
Bioavailability of capsule was
increased threefold by food
Diltiazem Extended 45mg SO 16 healthy male Cmax and AVC increased 37% FRISHMAN
release capsule subjects and 13 %, respectively (1993)
Encainide 35mg MO IS healthy male Standard breakfastC 120ml water Mid-meal 24h serum Bioavailability of encainide HILLEMAN
q8h subjects increased 1.8-fold and that of et a!. (1992)
O-desmethyl encainide 1.3-fold.
3-Methoxy-desmethyl
encainide unchanged, as were
electrocardiograms
Felodipine and Tablet 5mg SO 6 healthy male 250 ml grapefruit 250ml water With water 8 h plasma Felodipine and BAILEyet al.
nifedipine subjects juice or orange juice or juice dehydrofelodipine AVC (1991)
increased 2.5- and 1.7- fold by
10mg SO grapefruit juice. Nifedipine
and dehydronifedipine AVC
increased 1.4- and 1.2-fold.
Orange juice had negligible
effect. Felodipine had greater
effects on diastolic blood
pressure and heart rate when
taken with grapefruit juice
Table 5. Continued
Drug Dosage form Dose Dosage Subjects Food details Fluid volume Time interval Sampling Result of food administration Reference
regimen duration

Felodipine Tablet 5 mg SD 9 healthy male 200 ml grapefruit 200 ml water With water or 8h plasma Cmax increased 2.5- and 2.9- EDGAR et a1.
subjects juice of different juice fold and AUC increased 2.9- (1992)
strengths and 3.3-fold by increasing
strengths of grapefruit juice.
Levels of dehydro-felodipine
similarly increased
Fenretidine Capsule 300mg SD 13 healthy male High-fat breakfast' 200 mI water Immediately 72h plasma Fenretidine Cmax and DOOSE et aI.
subjects after meal AU C increased threefold, (1992b)
metabolite Cmaxand AVe
increased 2.3- and 2.5-fold
Fenretidine Capsule 300mg SD 15 healthy male High-fat, 200ml water lOmin after 72h plasma High-fat meal increased DOOSE et a1.
subjects -carbohydrate or meal bioavailability threefold (1992b)
-protein mealsc relative to carbohydrate meal.
High-protein meal
gave intermediate values
Gepirone Capsule 20mg SD 20 healthy male Standard breakfast 200 ml water 15 min after 48h plasma AVe increased lA-fold. Cmax TAyet al.
subjects meal slightly reduced and delayed (1993)
Itraconazole Capsule 100mg SD 6 healthy male Standard breakfast C 200ml water Immediately 96h plasma Cmax increased 3.4- VAN PEER
subjects after meal fold, AUC 2.6-fold et al. (1989)
Itraconazole Capsule 200mg SD 28 healthy male Standard breakfast C 200ml water Immediately 72h plasma Cm~ increased 2- fold, AUC BARONE et a1.
subjects after meal 1.6-fold. Metabolite Cm.. (1993)
increased l.4-fold, AUC
l.5-fold
Itraconazole and Capsules 100mg 12 healthy Standard light or full 100 ml water 5 min after 96 h plasma Itraconazole bioavailability ZIMMERMANN
fluconazole subjects, 7M, 5F breakfast' meal increased l.4-fold by light et aI. (1994)
for each drug meal, and 1.7-fold by heavy
meal. Fluconazole unaffected
Levodopa Immediate or 100mg IR SD 5 healthy male Light breakfast After meal 8h plasma Absolute bioavailability WILDING et a1.
controlled 200mg CR subjects relative to intravenous dose: (1991b)
release tablet IR fed> IR fasted> CR fed
> CR fasted. Food increased
absorption from both
formulations
Levodopa Variable MD 16 patients with Diet rich in soluble Few min after 6h plasma Total plasma levodopa levels ASTARLOA
idiopathic fiber (DRIF)' fiber supplement increased l.3-fold, levels et al. (1992)
Parkinson's during first h postdose
disease and increased 1.5-fold, after
severe 2 weeks on D RIF
constipation,
6M,13F
5-Methoxy- Tablet 20mg SD 9 healthy Standard breakfast' 100ml water 20-70 min after 10 h plasma C~, ranged from 0 to 88 ng/ml EHRSSON et a1.
psoraien volunteers, meal fasting and 37~141 ng/ml (1994)
IM,8F nonfasting. AVC ranged from
oto 422 ng·h/ml nonfasting
Moclobemide 100mg SD 8 healthy male Standard high- 9 h plasma Trend toward higher absolute A.F.D. COLE
subjects protein breakfastc bioavailability after protein et a1. (1992)
meal (67%) compared to
fasting (58%)
Nifedipine Extended 40mg 20 healthy male Low-fat and high-fat 45 min after 48h blood C max increased 1.8-fold and KLEINBLOESEM
release tablet subjects meal meal 2.4-fold, AVC increased 1.2- et a1. (1993)
fold and 1.2-fold, by low-fat
and high-fat meals,
respectively
Oxcarbazepine Tablet 600mg SD 6 healthy male Standardized high- 100 ml water Immediately 72 h plasma Cm~ and AVC of parent drug DEGEN et al.
subjects fat, high-protein after meal increased 1.7- and 1.2-fold, (1994)
breakfast respectively. Values for
monohydroxy and dihydroxy
metabolites also increased
Oxybutinin Solution 15 mglkg SD 18 healthy male High-fat breakfast With meal Bioavailability YONG et al.
subjects increased 1.3-fold (1991)
Phenytoin Powder 5 mglkg 4 healthy male Low-fat and high-fat 200ml water Immediately 34h plasma Cmax increased 1.5- and 2.2- HAMAGUCHI
subjects breakfast' after meal fold by low-fat and high-fat et a1. (1993)
meals, respectively. AVe
increased 1.5- and 1.9-fold
Progesterone Capsule 200mg MD 15 healthy Standard breakfast' Immediately 24 h plasma On day 1, Cmaxincreased SIMON et a!.
5 days postmenopausal after or 2 h on day 1, 10 5-fold and AVe 2-fold. (1993)
women before h plasma on Similar effects on day 5
day 5
Repirinast Tablet 300mg SD 12 healthy male High-carbohydrate l00ml water Immediately 72 h plasma Cmu and A V e increased SCHAEFER
subjects or high-fat breakfast' after meal 1.6- and 1.9-fold by high- et a1. (1993)
carbohydrate breakfast and
3.2- and 2.4-fold
by high-fat breakfast
Sparfloxacin 300mg SD 3 healthy subjects ~ Cm~ increased 1.5-fold, AVC TANIMURA
1.2·fold, and urinary excretion et a1. (1991)
1.4-fold
S-1108 200mg SD 6 healthy male Japanese breakfast l00ml water Immediately lOh blood, Cm~ increased l.l-fold, AVe SAITO (1993)
subjcets -500 Kcal after meal 24 h urine 1.4-fold
S-1108 l00mg SD Healthy male C max increased l.4·fold, AVe NAKASHIMA
subjects 1.5·fold. T max increased from et a1. (1993)
1.4 to 2.5 h
Table 5. Continued
Drug Dosage form Dose Dosage Subjects Food details Fluid volume Time interval Sampling Result of food administration Reference
regimen duration

Theophylline Capsule 900mg SD 20 healthy High-fat breakfast' 15 min after 72 h serum Cmax increased 1.6-fold, COOK et al.
subjects meal AUC 1.2-fold (1990)
Ticlopidine Tablet 250mg SD 12 healthy male High-fat breakfast' 200ml water 30 min after 24h plasma Cm~andAUC SHAH et al.
subjects meal increased 1.2-fold (1990)
Tramadol Tablet 100mg SD 18 healthy male High-fat breakfast Immediately 36 h plasma C_ and AUC of tramadol LIAO et a1.
subjects after meal increased 1.2- and l.l-fold, (1992)
respectively. Metabolite levels
only moderately affected
Vanoxerine Tablet 100mg SD 12 healthy male High-fat or low-fat 100 ml water After meal C_ increased 1.5-fold by INGWERSEN
subjects breakfastC high-fat meal, unaffected et al. (1993)
by low-fat meal. AUC
increased 1.8-fold and 3.6-fold
by low-fat and high-fat
meals, respectively
Vinpocetine Tablet 40mg SD 8 healthy Standard breakfast, 150ml 10 min before 12 h serum Cmax increased 1.6-, 2.3-, and LOHMANN
subjects, 4M, 1350 Kcal mineral start of meal, 1.9-fold while AUC increased et al. (1992)
4F water 10 min after 1.6-,1.7-, and 2.0-fold when
start of meal, drug was taken 10 min before,
20 min after 10 min after starting, and 30
termination of min after finishing meal
meal
Zalospirone 10mg MD 2418- to 45- Medium-fat meal 30 min after Cmax and Ave increased KLAMERUS
q8h year-old and 24 meal twofold. Elderly subjects gave et al. (1993)
" 65-year- higher levels than younger
old male and subjects
female subjects
566C80 Tablet 500mg SD 18 healthy male 45 min after Cmax increased 5A-fold, AVe ROLAN et a1.
subjects meal 3.3-fold (1992)
566C80 Tablet 500mg SD 12 healthy male Toast, toast plus 28 emax increased 2.5- and 3.2- ROLAN et al.
subjects g butter (LOFAT), fold after LOFAT and (1992)
toast plus 56 g butter HIFAT, respectively,
(HIFAT) compared to plain toast.

a Data not available.

b Single dose.
C Not relevant.
d Meal described.
e MUltiple doses.
Drug-Food Interactions Affecting Drug Absorption 85

again, not only the broad spectrum of effects that may result from drug-food
interactions, but also their frequent unpredictability. The compounds in this
section tend to be lipophilic and poorly water soluble, but this is not always the
case. Similarly some interactions are relatively trivial and do not warrant
further discussion, while others are substantial and likely to be clinically
significant.
Amiodarone is an amphiphilic substance and its bioavailability is ex-
tremely variable (LATINI et al. 1984; RIVA et al. 1982). In order to examine its
absorption in the presence of food, amiodarone was administered into the
jejunum together with two nutrient solutions, one 3.3 Kcal/min and the other
1.3 Kcal/min, with the former containing total lipid and caloric load 2.5 times
greater than the latter (PFEIFFER et al. 1990). Based on intestinal testing,
absorption of amiodarone correlated significantly with lipid absorption rate.
However, plasma Cmax and AVC values were extremely variable and tended to
be higher from the 1.3-Kcal/min infusion. These seemingly paradoxical results
were attributed to wide fluctuations in amiodarone pharmacokinetics, distri-
bution, and metabolism.
Conflicting results from drug-food interactions have been reported with
the onchocerciasis agent amocarzine (CGP 6140) (LECAILLON et al. 1991). In
20 male Guatemalan patients who received a 3-mg/kg oral dose, systemic
availability was increased 20% when drug was taken with a "copious" break-
fast compared to fasting. In a similar study, when the dose was increased to
1200mg, both the peak plasma levels and systemic availability of amocarzine
were increased approximately threefold when drug was given after a meat and
noodle standard breakfast, relative to fasting (LECAILLON et al. 1990). The
AVC of the metabolite CGP 13231 also increased threefold in this study,
although it was unaffected at the lower dose. Mean plasma levels of
amocarzine and its metabolite CGP 13231 are shown in Fig. 11.
The considerable increase in absorption due to food after the high dose of
amocarzine may be related to a greater degree of solubilization after the meal
or to decreased presystemic metabolism. The nature of the food in the lower-
dose study was not described, but is likely to have been similar to that used in
the high-dose study. A controlled high-dose, low-dose study would provide
useful mechanistic information on this interaction. Substantially increased
absorption due to food has been reported for the lipophilic antiprotozoal
agent atovaquone (ROLAN et al. 1994). In healthy male volunteers, peak
atovaquone plasma· levels increased over fivefold while systemic
bioavailability increased over threefold when drug was given 45 min after a
Western-style high-fat breakfast compared to fasting. Mean plasma profiles
obtained over 48 and 528h are shown in Fig. 12.
Complementary studies using meals with different fat content, aqueous
suspensions, and oily emulsion vehicles, and also pre dosing with a cholecysto-
kinin octapeptide, led to the conclusion that the observed food effect with
atovaquone was probably due to the combined effects of bile release and also
increased solubility due to a direct effect of the fatty meal.
86 P.G. WELLING

15000
12000 A

9000
6000

S0 3000
E
E
c 6 12 18 24
.
0
~
C
~
15000
<oJ
C
0 12000 B
u
9000
6000
3000

I
6 12 18 24
Time (hr)
Fig.llA,B. Mean plasma versus time concentrations (±SD) of CGP 6140 (A) and its
N-oxide metabolite CGP 13231 (B) in 11 patients following a single 1200-mg oral dose
of CGP 6140 fasting (0) or following a large breakfast (e). (From LECAILLON et al.
1990)

More modest drug-food interactions are reported for the 5-lipoxygenase

inhibitor BAY-X-1005 (BECKERMANN et al. 1993), the selected monoamine
oxidase-A inhibitor brofaromine (DEGEN et al. 1993), and the vasoactive agent
buflomedil (WILSON et al. 1991). With all three compounds, administration
after a standard breakfast resulted in 1.2-fold to 2-fold increases in absorption.
In the case of buflomedil, the extent of absorption was increased 1.2-fold after
a heavy meal compared to a light meal, while increases for the other agents
were observed relative to the fasting state.
Consistent with some previous observations on other cephalosporins
(WELLING 1977), food has been shown to delay but also to moderately increase
absorption of cefetamet pivoxil. Absorption of cefetamet pivoxil, the
pivaloyloxymethyl ester prodrug of cefetamet, was delayed by food, and mean
peak plasma levels occurred at 4.Sh compared to 3h fasting. Overall
bioavailability and peak plasma levels increased approximately 25%-30%
(BLOUIN and STOECKEL 1993). A similar effect to this was observed when
cefetamet pivoxil was administered 1 h after a standard breakfast, although
plasma profiles were similar when drug was administered either with or 1 h
before the standard breakfast (TAM et al. 1990).
Increasing fluid volume had no effect on absorption. Plasma profiles ob-
Drug-Food Interactions Affecting Drug Absorption 87

4
E
~
~
cu 3 0
c::
0 84 168 252 336 432 528
j
cr

-
t':S
:>
0
t':S 2
t':S
E
<I>
t':S
'2i:
1

o~--------~--------~~--------~--------~
o 12 24 36 48
Time (hr)

Fig. 12. Mean plasma versus time concentrations of atovaquone in 18 subjects follow-
ing a single SOO-mg oral dose of atovaquone fasted (0) or 4Smin after a high-fat meal
(e). (From ROLAN et al. 1994)

tained with cefetamet pivoxil administered under fasting and nonfasting con-
ditions, and with different fluid volumes, are shown in Fig. 13. The lack of
effect of fluid volume in this study may have been due to the relatively large
fluid volumes used, 250ml and 450ml. However, the lack of a food effect when
drug was administered directly with the meal was unexpected. A proposed
explanation is that while drug was taken with water 1 h before and 1 h after
breakfast, it was taken with the tea or coffee provided for the group that
received the drug with food. Previous studies have shown that bioavailability
of cefuroxime is increased when taken with food (WILLIAMS and HARDING
1984). In an attempt to identify the mechanism of this interaction, cefuroxime
was administered to healthy male subjects 30 min before intravenous hyoscine
butylbromide or immediately before intravenous cholecystokinin 8 (MAcKAY
et al. 1991). Hyoscine butylbromide had no effect on cefuroxime absorption
while cholecystokinin resulted in a 20% increase in cefuroxime emax and AVe
values. These results led to the conclusion that bile release, but not gastric
emptying, may be at least partially responsible for increased cefuroxime ab-
sorption in the presence of food.
Possibly the most remarkable food effect reported during the review
period involved the lipophilic hypolipidemic compound eGP 43371 (SUN et al.
1994). Administration of single 800-mg capsule doses of eGP 43371 after a
88 P.G. WELLING

7
6 Cl"',Ci
~

E 5 .'0' ......
II

-
II
B
~ I
\\
\\

....
\\
4
-
Q,I \\
E
~
\\
\\

JQ,I! 3
\J
51...
'\
~
E 2
<II
\~
~
ii: .... ....
1
0
0 2 4 6 8 10 12
Time (hr)
Fig. 13. Mean plasma versus time concentrations of cefetamet in 12 subjects following
a single lOoo-mg oral dose of cefetamet pivoxil under fasted (-) or fed (---) conditions
with 250ml (0) or 450ml (0) water. (From TAM et al. 1990)

2.5

1 2.0
~
,...
I'-. 1.5
..,
M
M

=- 1.0
1..7
u
~
E
<II
~
ii: 0.5

0.0
0 12 24 36 48 60 72 84 96
Time (hr)
Fig.14. Mean plasma versus time concentrations of CGP 43371 in 12 subjects following
a single Boo-mg oral dose of CGP 43371 as a dispersion (6) or capsule (0) under fasting
conditions or as a capsule after a standard meal (e). (From SUN et al. 1994)

standard breakfast caused an ll-fold increase in peak plasma drug levels and
a 13-fold increase in overall bioavailability relative to the fasting state. Plasma
levels from this study are shown in Fig. 14.
It is proposed that, as CGP 43371 is absorbed mainly from the ileum,
delayed gastric emptying would enable more compound to disintegrate and
Drug-Food Interactions Affecting Drug Absorption 89

dissolve before reaching this absorption site. It is further proposed that CGP
43371 dosage should be modified depending on dosing relative to food intake.
This study provides a truly dramatic example of the extent to which drug-food
interactions can influence circulating drug profiles, with obvious clinical
implications.
Similar food effects were observed in infants and adults for the new
macrolide antibiotic clarithromycin (CHU et al. 1992; GAN et al. 1992). Follow-
ing a 7.5-mg/kg dose to infants and children aged 6 months to 10 years with
various infections, either fasting or after milk and/or hash brown potatoes,
peak plasma levels were 4.6,ug/ml and 3.6,ug/ml after nonfasting and fasting
doses, respectively (GAN et al. 1992). Although these increases were modest,
systemic availability increased by 40%, indicating better overall absorption
with food. In the study in adults, food taken immediately before a 500-mg
clarithromycin dose increased the extent of absorption by approximately 25%
(CHU et al. 1992). In both of these studies plasma levels of the major active
metabolite 14-hydroxyclarithromycin were moderately increased by approxi-
mately 10%. Different results were obtained from blood and plasma analysis
regarding the effect of low-fat and high-fat meals on cyclosporine absorption
(GUPTA et al. 1990). Based on plasma data, cyclosporine systemic availability
was 23% and 42% after low-fat and high-fat meals, respectively. Based on
blood data, relative values were 21 % and 79%. Thus, the apparent increase in
cyclosporine bioavailability appears to depend on the sample matrix. This in
turn is probably related to differential penetration of cyclosporine into red
cells depending on the amount of absorbed fat. Plasma to blood cyclosporine
clearance ratios were 1.8 and 1.3 after high-fat and low-fat meals, respectively.
Other studies have shown that high-fat meals increased cyclosporine
clearance, but not mean residence time, based on either blood or plasma
measurements (GUPTA and BENET 1990).
Absorption of the heterocyclic steroid derivative danazol (CHARMAN et al.
1993) and also the retinoid fenretinide (DOOSE et al. 1992b) is substantially
increased by food. In the case of danazol, systemic availability from a capsule
dose was increased over threefold by food in healthy female subjects, resulting
in mean peak plasma levels of 37 and 101 ng/ml, after fasting and fed doses,
respectively. In the case of fenretidine, bioavailability and peak plasma levels
both increased threefold following a high-fat meal compared to fasting (DOOSE
et al. 1992). Administration of fenretidine in an oil suspension to fasting
subjects yielded intermediate values. Further examination of the effect of meal
composition showed that a high-fat meal resulted in plasma fenretinide
bioavailability three times greater than a carbohydrate meal, with a high-
protein meal yielding intermediate results. Mean plasma profiles obtained in
these studies are shown in Figs. 15 and 16. These results give rise to the
recommendation that fenretinide be given with food in order to optimize
potential therapeutic benefit.
A specific drug-food interaction has recently been identified between a
component in grapefruit juice and the calcium antagonist felodipine (BAILEY
90 P.G. WELLING

600

500
E
~
-=-
~
400
~
·2
;:; 300
~
i:
~
til 200
E
'"til
E: 100

0
0 12 24 36 48 60 72
Time (hr)

Fig.1S. Mean plasma versus time concentrations offenretinide in 13 subjects following

a single 300-mg oral dose of fenretidine fasting (0), after a meal (e), and as a 20-ml
neutral oil suspension administered fasting (L). (From DOOSE et al. 1992b)

600

500
~
-=-
~
400
·c
~

.
;:;
~

c
300

~
til 200
E
'"til
E: 100

0
0 12 24 36 48 60 72
Time (hr)

Fig.16. Mean plasma versus time concentrations of fenretinide in 15 subjects following

a single 300-mg oral dose of fenretidine after a high-fat (e), high-protein (0), or high-
carbohydrate (.~) meal. (From DOOSE et al. 1992b)

et al. 1991; EDGAR et al. 1992). In a study in six men with borderline hyperten-
sion, felodipine and dehydrofelodipine systemic availability increased 2.5- and
1.7-fold, respectively, when felodipine was taken with two 2S0-ml double-
strength grapefruit juices, relative to water (BAILEY et al. 1991). Under the
same conditions, plasma levels of nifedipine and dehydronorfedipine in-
creased 1.4- and 1.2-fold. The results with felodipine were reproduced in
another study in nine healthy middle-aged men (EDGAR et al. 1992). The
interaction with grapefruit juice, which is believed to be a class effect for the
Drug-Food Interactions Affecting Drug Absorption 91

dihydropyridines, is thought to be due to inhibition of first-pass oxidative

metabolism by flavonoids in the grapefruit juice. However, the precise mecha-
nism of interaction has yet to be identified.
Food has been shown to increase the systemic availability of the antifungal
agent itraconazole under a variety of conditions. Peak plasma itraconazole
levels increased 3.4-fold, and systemic availability 2.6-fold, in six healthy male
subjects following a standard breakfast compared to fasting (VAN PEER et al.
1989). Values for these parameters were increased 2- and 1.6-fold following a
standard breakfast in an expanded study in 28 male volunteers (BARONE et al.
1993). In contrast to itraconazole, absorption of the closely related compound
fluconazole was relatively insensitive to food, both ernax and AVe being
slightly reduced by a full meal and essentially unchanged by a light meal,
relative to fasting. While these divergent results are consistent with previous
data on these agents, there is no mechanistic explanation for their different
behavior. Food effects on absorption of levodopa were influenced by the type
of meal but to a lesser extent by the dosage form. In six healthy volunteers
levodopa absolute bioavailability from an immediate release dosage form was
86.4% and 80.4% from fed and fasting treatments, respectively. Levodopa
availability from a controlled release dosage form was 71 % and 63.6% from
fed and fasted treatments (WILDING et al. 1991b). Thus, although the con-
trolled release dosage form yielded lower absolute bioavailability, the food
effect was similar for both formulations, increasing bioavailability by approxi-
mately 10%. On the other hand, a diet rich in insoluble fiber (DRIF) was
shown to increase levodopa plasma levels by 30%, and levels during the 1st h
postdose by 50%, in patients after 2 weeks on a DRIF diet compared to
baseline (ASTARLOA et al. 1992). Thus, the DRIF may serve a useful dual
purpose of relieving constipation and also increasing plasma levels and pre-
sumably effectiveness of levodopa. As the effect on levodopa levels occurred
mainly during early periods after drug administration, it appears that levodopa
absorption was accelerated due to increased gastrointestinal motility and
shorter gastric emptying time.
While the effect of food on nifedipine absorption from conventional tab-
lets is somewhat variable, food appears to give rise to a moderate but signifi-
cant increase in nifedipine absorption from an extended release tablet
(KLEINBLOESEM et al. 1993). Peak plasma nifedipine concentrations were in-
creased 1.8- and 2.4-fold by low-fat and high-fat meals, respectively. Overall
systemic availability was increased 1.2-fold by both treatments. Increased
nifedipine plasma levels appeared to be without effect on blood pressure
relative to the fasting state, but mean heart rate increased by 10bpm after both
postprandial doses compared to 5 bpm fasting. Peak plasma levels and sys-
temic availability of the major monohydroxy metabolite of the antiepileptic
compound oxcarbazepine were increased 1.2- and 1.5-fold when a 600-mg
tablet dose was taken after a high-fat, high-protein breakfast (DEGEN et al.
1994). Plasma concentrations of the dihydroxy metabolite were similarly af-
fected by food intake. Powders of different particle size have been shown to
92 P.G. WELLING

give rise to different food-drug interactions for the anticonvulsant agent

phenytoin. Using a commercial Japanese powder of mean particle size 190 J1M,
systemic availability was increased 1.5- and 2.2-fold by low-fat and high-fat
meals (HAMAGUCHI et al. 1993). Plasma profiles in four subjects are shown in
Fig. 17. The marked increase in plasma phenytoin levels due to food in this
study differs from results obtained earlier by SEKIKAWA et al. (1980), who used
a Japanese commercial powder of mean particle size 47.1J1M. The different
results obtained in the two studies are probably related to slower dissolution
from the larger particle size formulation in the fasted state and a greater,
positive effect on the systemic availability from this formulation due to the
direct influence of food and also delayed gastric emptying. A far more dra-
matic food effect occurred with oral micronized progesterone (SIMON et al.
1993). Repeated doses of micronized progesterone were administered in cap-
sules to 15 healthy postmenopausal women, either 2h before or immediately
after a standard breakfast for 5 days. Plasma profiles were obtained on days 1
and 5. On day 1, peak plasma levels of progesterone, obtained by radioimmu-
noassay, were increased fivefold and systemic availability twofold by food.
Similar results were obtained on day 5. Increased progesterone absorption due
to food was attributed either to a direct drug-food interaction in the gas-
trointestinal tract or to increased blood flow to the liver causing decreased
pre systemic clearance. No attempt was made to further examine or differenti-
ate these possible mechanisms. A further example of differential food effects

o~--------~----------------~------~
o 10 26 34
Time (hr)

Fig. 17. Mean plasma versus time concentrations (±SE) of phenytoin in four subjects
following a single 5-mglkg oral dose of free acid phenytoin powder with large particle
size fasting (0), and after low-fat (e) and high-fat (.) meals. (From HAMAGUCHI et al.
1993)
Drug-Food Interactions Affecting Drug Absorption 93

on the extent of drug-food interaction is provided by the antiallergy agent

repirinast (SCHAEFER et al. 1993). In a study conducted in 12 healthy men,
peak plasma levels and systemic availability of the active de-esterified metabo-
lite were increased 1.6- and 1.9-fold by a high-carbohydrate continental break-
fast, and 3.2- and 2.4-fold by a high-fat American breakfast. The terminal
elimination half-life of the active metabolite from plasma was considerably
prolonged by both meals (12.7h vs 2.5h for the high-carbohydrate meal and
15.4h vs 7.1 h for the high-fat meal), suggesting prolonged absorption in the
nonfasting state.
Similar to observations with cefetamet pivoxil (BLOUIN and STOECKEL
1993; TAM et al. 1990), food had a positive effect on the new ester-type oral
cephalosporin S-llOS. In two separate studies conducted in Japan the systemic
availability of S-llOS was increased approximately 1.5-fold, and peak plasma
levels 1.1- to l.4-fold, following a Japanese-style breakfast (NAKASHIMA et al.
1993; SAITO 1993). Urinary recovery was increased by both treatments in both
studies. S-llOS was also administered to patients fasting lOmin before intra-
muscular ceruletide diethylamine, or 90min after oral ranitidine hydrochlo-
ride. Ceruletide diethylamine had no effect on S-llOS absorption, but Tmax was
delayed. Ranitidine, on the other hand, had a negative effect on S-llOS ab-
sorption. Thus, neither increased bile flow nor increased gastric pH seemed to
contribute to any food-related increase in S-llOS absorption. Delayed
gastric emptying appears to be a reasonable alternative.
Further studies with theophylline demonstrated increased absorption due
to food from single oral doses of Theo-24 capsules to healthy subjects (COOK
et al. 1990). Mean peak plasma theophylline levels increased 1.6-fold and
systemic availability 1.2-fold when this theophylline formulation was taken
immediately after a high-fat breakfast compared to the fasting state. Addi-
tional experiments comparing dog and man showed that absorption of theo-
phylline in the two species followed the same pattern regarding food effects
and also for different theophylline formulations, thus providing support, for
these theophylline products and study conditions at least, for predictability of
dog bioavailability data in comparison to man for commercial oral formula-
tions. Modest food-related increases have been reported for the platelet in-
hibitor ticlopidine (SHAH et al. 1990) and the analgesic agent tramadol (LIAO et
al. 1992). A more dramatic interaction was observed with the piperazine
derivative dopamine reuptake inhibitor vanoxerine (INGWERSEN et al. 1993).
Administration of vanoxerine as a 100-mg dose in 4 x 25-mg tablets to 12
healthy men after low-fat and high-fat breakfasts increased systemic availabil-
ity 1.S-fold and 3.6-fold, respectively. Despite the considerable increase in
systemic availability after the high-fat meal, emax was increased less than
twofold due to delayed absorption. Peak drug plasma levels occurred at 2.5 h
after the high-fat meal compared to 45 min for fasting. The mechanism of
increased absorption in this study was not addressed. However, one subject
who was virtually unaffected by food intake was a "poor metabolizer" of
debrisoquine, which suggested that decreased first-pass metabolism, possibly
94 P.G. WELLING

related to increased splanchnic blood flow, may have contributed to the food
effect in the other subjects.
Absorption of the nootropic agent vinpocetine and also the 5-HT1• partial
antagonist zalospirone is modestly increased by food (LOHMANN et al. 1992;
KLAMERUS et al. 1993). Administration of vinpocetine tablets 10min before
and 10min after starting and 30min after a standard 1350-Kcal breakfast
increased systemic availability 1.6-, 1.7-, and 2.0-fold relative to fasting. Peak
plasma levels were affected similarly. Mean plasma profiles in eight healthy
male and female subjects in this study are shown in Fig. 18. Peak plasma levels
and areas under plasma curves of zalospirone were increased approximately
lA-fold by food in both young (18-45 years old) and elderly (~65 years old)
subjects (KLAMERUS et al. 1993). In addition, plasma levels were almost
doubled in elderly subjects relative to young subjects. The results of these
studies led to the recommendation that both vinpocetine and zalospirine be
taken with or after meals.
The last drug considered in this category illustrates again the dramatic
positive effect that food can have on circulating drug profiles. In a study in 18
healthy men, the systemic availability of a novel anti protozoal agent 566C80
was increased 3.3-fold, while e m• x was increased SA-fold when administered

30

E
~ 20
5
.c
.;:
ClJ
U
0
c..
c
.;;:
!U
E
<n
!U 10
E:

2 3 4 5 6
TIME (hr)
Fig. 18. Mean plasma versus time concentrations of vinpocetine in eight subjects
following a single lO-mg oral dose of vinpocetine fasting (0) and lOmin before (1:.),
10min after (e), and 30 min after (.) starting a standard breakfast. (From LOHMANN et
al. 1992)
Drug-Food Interactions Affecting Drug Absorption 95

after food (HUGHES et al. 1991). In attempts to elucidate the mechanism of this
interaction, 566C80 was given fasted and with meals of varying fat content, as
an aqueous suspension, and as a oily emulsion. In a related study 566C80 was
given after an infusion of cholecystokinin octapeptide (CCK-OP) (ROLAN et
al. 1992). Pharmacokinetic results from these studies led to the conclusion that
increased absorption of 566C80 after food could be quantitatively accounted
for by dietary fat and that stimulation of bile secretion may be a small compo-
nent of the food effect.

G. Interactions Causing Accelerated Drug Absorption

Although, as described in the previous section, food may often increase drug
absorption and this may be accompanied by faster absorption rates, there have
been few reports of food increasing the rate of drug absorption but having no
effect on overall systemic bioavailability. Three examples of this phenomenon
are shown in Table 6 and are described here.
The nonacidic, nonsteroidal anti-inflammatory agent nabumetone is con-
verted in the liver to the active metabolite 6-methoxy-2-naphthylacetic acid
(6MNA). This is the predominant form circulating in plasma. Administration
of nabumetone to healthy individuals with food resulted in a 1.5-fold increase
in peak 6MNA plasma levels, but overall systemic availability was unaffected,
relative to fasting (HYNECK 1992). Administration of nabumetone with milk or
antacid had a similar effect on 6MNA plasma profiles. The lipophilic character
of nabumetone may account for increased absorption rates in the presence of
food and milk, but an alternative explanation is necessary to explain the
antacid effect. Increased dissolution is unlikely for this non acidic molecule.
6MNA profiles in fasted individuals and the effects of food, milk, and antacid
are shown in Fig. 19.
Increased absorption rate of naproxen with food, resulting in a modest
reduction in Tm • x and a greater increase in the 0- to 2-h area under the
naproxen plasma curve, may have been due, in part at least, to a fluid volume
effect (LAFONTAINE et al. 1990). The treatment scheme for naproxen with food
in healthy subjects included 675ml drinking water, while the fasted treatments
provided only 300ml water. Ingested food may nonetheless have contributed
to faster naproxen absorption causing a transient increase in gastric pH,
thereby increasing the dissolution rate of naproxen, which is poorly soluble in
acidic media. Increased biliary excretion may have also contributed to acceler-
ated naproxen dissolution and absorption. Although absorption of valproate
from conventional tablets is delayed by food (OHDO et al. 1992), absorption
from a sustained release formulation containing a mixture of sodium valproate
and valproic acid appears to be accelerated (ROYER-MoRROT et al. 1993). In a
study using a Depakine Chrono 500-mg sustained release valproate/valproic
 1.0
0\

Table 6. Drug food interactions resulting in accelerated drug absorption

Drug Dosage Dose Dosage Subjects Food details Fluid Time Sampling Result of food Reference
form regimen volume interval duration administration
_b 12 healthy High-fat meal 96h plasma
Nabumetone lOOOmg SD' - Cm~ increased HYNECK
subjects 1.5-fold, AUe (1992)
unchanged

Naproxen Tablet 500mg SD 6 healthy Low-fat meal' - At beginning 48h plasma AUC(O-2h) LAFONTAINE
subjects of meal increased. Fluid et al. (1990)
volume may have
contributed

Valproate sodium Sustained 333mg valproate SD 12 healthy Standard 30ml water Midpoint of 72h plasma Cm~ and 6h ROYER-MoRROTT
and valproic acid release sodium, 145 mg female breakfast' meal absorption et al. (1993)
formulation valproic acid subjects increased 1.2-fold.
Systemic
availability
unchanged

• Single dose.
b Data not available.
, Meal described.

'"
o
~
E
z
Cl
Drug-Food Interactions Affecting Drug Absorption 97

50

40

<:
z 30
~
..b
os
E
'I>
os 20
E:

10

0
0 10 20 30 40 50 60 70 80 90 100
TIME (hr)

40

B
<: 30
z
~
..b
os
E 20
'I>
os
E:

10

0
0 12 24 36 48 60 72

TIME (hr)

Fig. 19A,B. Mean plasma versus time concentrations of the active metabolite of
nabumetone in A 12 subjects following a single lOOO-mg oral dose of nabumetone
fasted (0) or after a meal (e), and in B 15 patients following a single lOOO-mg oral dose
of nabumetone administered with water (0), milk (e), and aluminum hydroxide (.).
(From HYNECK 1992).

acid formulation in healthy young women, ingestion at the mid-point of a

standard breakfast caused a 1.2-fold increase in both the peak plasma
valproate levels and 6h absorption, while overall systemic availability was
unchanged, relative to fasting. While increased bile flow and faster dissolution,
may have contributed to the faster absorption rate, the divergent results
obtained in the two studies suggest that the latter result may be a formulation
effect rather than a direct drug effect.
98 P.G. WELLING

H. Cases in Which Food Has No Effect

on Drug Absorption
The reports summarized in Table 7 describe studies in which food had no
effect, or an insignificant effect, on the rate and extent of drug absorption. In
some cases compounds appearing in this table have already been cited under
previous categories in Tables 3-6 and this again reflects the varied results that
may be obtained under different study conditions, from different laboratories,
or from different formulations of the same drug. Many of the results described
in Table 7 are results obtained from routine screening for drug-food interac-
tions, while others arise from deliberate strategies to minimize the influence of
factors such as co administered food on the rate and extent of drug absorption.
This may have been the case with the triazolobenzodiazepine alprazolam,
whose absorption was essentially unchanged by food when it was administered
in a prototype mixed polymeric controlled release tablet formulation. Mean
peak circulating drug levels increased by 12 % with food, but other pharmaco-
kinetic parameters were unchanged (ELLER and DELLA-COLETTA 1990). Lack
of food effect in this case is not surprising as in vitro dissolution rate for this
formulation was shown to be pH-independent and in vivo plasma clearance of
alprazolam is low so that metabolism is primarily determined by hepatic
metabolic capacity rather than by blood flow. In the case of the new
dihydropyridine calcium channel antagonist amlodipine, plasma profiles of
unchanged drug in fasting and fed subjects were identical, as were profiles
from capsule and solution doses. Plasma profiles obtained in this study are
shown in Fig. 20. In the case of bambuterol, food had no effect on circulating
levels of either bambuterol or its active metabolite terbutaline (ROSENBORG
et al. 1991).
While the absorption of hydrochlorothiazide has previously been reported
to be both increased (BEERMAN and GROSCHINSKy-GRIND 1978) and decreased
(BARBHAIYA et al. 1982) by food from conventional single-drug formulations,
absorption of both hydrochlorthiazide and bisoprolol was unaffected by food
when these drugs were administered in a combination tablet to healthy volun-
teers (MURALIDHARAN et al. 1992). No significant differences were observed in
plasma pharmacokinetic parameters, except that the hydrochlorothiazide Tmax
was increased 26%, or in the percentage of hydrochlorothiazide excreted in
urine.
Improved safety of the selective monoamine oxidase A inhibitors with
respect to the classical "cheese" effect is demonstrated by the compound
brofaramine. Ingestion of brofaramine together with cheese containing the
equivalent of the PD 30 dose of tyramine resulted in no change in blood
pressure in one subject, and a maximum change of only 20mmHg in three
subjects. The mean increase in blood pressure was only 11 mmHg compared
to 40 mmHg from an equivalent dose of tyramine. It is claimed that the lack
of interaction with tyramine-rich foods greatly increases the benefit-risk
ratio of these MAO-A inhibitors (BlECK et al. 1993). Neither food nor
Table 7. Cases in which food had no effect on drug absorption
Orug Dosage from Dose Dosage Subjects Food details Fluid volume Time interval Sampling Result of food Reference
regimen duration administration

Alprazolam Matrix SR Img SO, 21 healthy Standard 170ml water 30min after 36h plasma Cmax increased 12 %. ELLERand DELLA-
tablet subjects, high-fat starting meal AUC and tm~ COLETIA (1990)
12M,9F breakfast b unchanged
Amlodipine Capsule lOmg SO 12 healthy male Standard 150ml water Within 30 min 192h plasma No effect on plasma FAULKNER et a1.
subjects breakfastb after completing levels (1989)
meal
Bambuterol 20mg MO'1 22 healthy Dinner Immediately No effect on plasma ROSENBORG
tablet each subjects after meal levels or urinary et al. (1991)
evening, excretion of
14 days bambuterol or active
metabolite terbutaline
Bisoprolol and Combination 10 mg bisoprolol, SO 16 healthy With meal 48h plasma Tm.. for MURALIDHARAN
hydrochlorothiazide tablet 6.25mg subjects and urine hydrochlorothiazide et al. (1992)
hydrochlorothiazide increased 20%
otherwise no effect on
plasma or urine
parameters
Brofaramine 150mg/day, 14 days MO 10 subjects Cheese or 25-75 mg - Blood Tyramine caused 40 BlECK et aL
tyramine pressure mm increase in blood (1993)
determination pressure. Mean
increase only 11 mm
with cheese
Bromocriptine Tablet 7.5mg SO 7 healthy male Standard breakfastb 100ml water After meal 14h plasma Cmax decreased 10%, KOPITAR et a1.
subjects AUC unchanged (1991)
Carbamazepine Tablet 200mg SO 12 patients with Very low protein AUC decreased 10%, BANNWARTH
mild to advanced diet Tmax reduced from et al. (1992)
chronic renal 8.5 to 6 h, emu
failure unaffected
Cardizem Capsule 360mg SO 18 healthy 36h plasma Diltiazem emax Yu et al. (1992)
subjects reduced 12%,
AU C unchanged.
Deacetyldiltiazem
Cmax increased 11 %,
AU C unchanged
Cefetamet pivoxil Syrup 385mg SO 18 healthy male Standard breakfast' 10ml syrup With meal 15h plasma, Plasma and urine DUCHARME
subjects 24h urine parameters for syrup et al. (1993)
unaffected.
Bioavailability from
syrup lower than
from tablet
Table 7. Continued
Orug Dosage from Dose Dosage Subjects Food details Fluid volume Time interval Sampling Result of food Reference
regimen duration administration

Cimetitine and Tablet 400 mg cimetidine SO 18 healthy male Standard breakfastb 100ml water 15min after 8h plasma No significant effect DESMOND et a1.
ranitidine 150 mg ranitidine subjects starting meal on plasma levels of (1990)
either compound. In
fasted subjects antacid
treatment reduced
plasma levels of
cimetidine and
ranitidine. In fed
subjects antacid had
no effect on either
drug
Cyclosporine Solution 3.51 ± 1.35 MO 14 renal Moderate and 125ml Immediately 12h plasma Cm"andAUC HONCHARIK
mg/kg/day (bj.d.) transplant trace-fat breakfastb orange juice after meal unaffected by trace- et aJ. (1991)
recipients, 9M, fat and moderate-fat
5F meals. Tmax prolonged
from 3.3 to 5.1 h by
moderate-fat meal
Cyclosporine Capsules -6mg/kg SO 11 healthy male Standard light 150ml water With meal 32h plasma Systemic availability LINDHOLM et aI.
subjects breakfast'. also bile unaffected by food (1990)
acid tablet but Cmax reduced
17%. Addition of
bile acid tablets
increased availability
25%
Diazepam, ethinyl 5mg SO 8-10 healthy Olestra or 6 ounces With diet Plasma up to No significant ROBERTS and LETT
estradiol, 0.07mg subjects triglyceride onb water 48h differences in (1989)
norethindrone, 1.0mg absorption when
propranolol 20mg administered with
olestra, triglyceride
oil, or water, except
Tmax for diazepam
prolonged with
triglyceride oil
Diazepam Conventional 60mgCR SO 24 healthy male Standard breakfast 200ml water 30 min after 24h plasma Food had no effect Du SOUICH et a!.
tablet and 120mgSR subjects starting meal on plasma levels of (1990)
slowwrelease diazepam and its
capsule metabolites from
either formulation,
but SR formulation,
increased systemic
availabilty 69%
relative to
conventional tablet
E2020 2mg SD 12 healthy male Standard breakfast Within 30min 168h plasma Rate and extent of MIHARA et a1.
subjects after meal absorption unaffected (1993)
Fluvoxamine Tablet SOmg SD 12 healthy Standard breakfastb 240ml water 15 min after 72h plasma Rate and extent of VAN HARTEN
subjects, 8M, 4F or orange starting meal absorption unaffected et a1. (1991)
juice
Ibuprofen 600mg SD 11 healthy male Standard breakfastb 200mi water 30 min after lOh plasma Plasma levels of S(+) LEVINE et a1.
subjects starting meal ibufrofen higher than (1992)
those of R(-)
ibuprofen under
fasting and nonfasting
conditions. Food
modestly reduced
Cm" of S(+) and
R(-), but systemic
availability, and
isomer ratios
unaffected
Levodopa Solution 125mg SD 8 healthy male 10.5 g or 30.5 g 100ml water 15 min after 8h plasma Meal with 30.5 g ROBERTSON
subjects protein mealsb meal protein had no effect et a1. (1991)
on levodopa
absorption,
Absorption was
reduced 10%, and
C_ by 26% by low-
protein meal
Methotrexate 9.S ± 5.8 mg/week SD 11 patients with Standard meal 24h plasma Rate and extent of PHELAN et a1.
rheumatoid absorption unaffected (1991)
arthritis
Metoprolol Controlled 400mg SD 18 healthy male Standard high-fat Tendency to increase SANDBERG et a1.
succinate release tablet subjects breakfast Cm" and AUC, but (1990)
not significant.
Similar emax values
Metoprolol Controlled 50mg SD 12 healthy Standard Plasma levels SANDBERG et a1.
succinate release tablet subjects carbohydrate unaffected by food (1988)
breakfastb
Morphine sulfate Sustained 30mg 9.12h MD 24 healthy male Standard caffeine- 180ml water Immediately 84hplasma Rate and extent of BASS et al. (1992)
release tablet subjects free meals after meals absorption unaffected
Mosapride citrate Tablet lOmg SD 10 healthy male Sandwiches and 200ml water 30 min after 24h plasma C~. and AUC not SAKASHITA et a1.
subjects 200 ml orange meal significantly affected. (1993)
juice Tmax increased from
0.6 to 0.9h
Moxonidine Tablet O.2mg SD 18 healthy male Standard breakfast 200ml water 30 min after 24h plasma Crnax decreased THEODOR et a1.
subjects meal and urine 14%, Trnax (1992)
marginally increased.
Systemic availability
and urinary excretion
unchanged
Table 7. Continued
Drug Dosage form Dose Dosage Subjects Food details Fluid volume Time interval Sampling Result of food Reference
regimen duration administration

Moxonidine Tablet 0.2mg SD 15 healthy male Cmax decreased WEIMANN and

subjects 14%, Tm " RUDOLPH (1992)
marginally increased.
Systemic availability
unchanged
Nefiracetam Tablet 100mg SD 39 healthy Standard light lOOml water 30min after 24h plasma Rate and extent of FUJIMAKI et a1.
Japanese male breakfast meal and urine absorption not (1992)
subjects affected
Paroxetine 30mg SD 41 healthy male Standard Immediately Rate and extent of GREB et al. (1989)
subjects continental after meal absorption not
breakfast, low-fat affected by standard
or high-fat diet, or breakfast, nor
milk between low-fat
meals. AUC reduced
42% by 1000ml
milk
Piroximone 25mg SD 12 healthy male Standard breakfastb - Immediately 24h plasma Following the 25-mg HAEGELE et a1.
or subjects before meal and urine dose, rate and extent of (1991)
50mg absorption, unaffected.
Following the 50-mg
dose, extent of
absorption marginally
increased
Procainamide Slow release SD Healthy subjects High-fat meal During meal Negligible effect on DEVRIES et al.
tablet rate and extent of (1991)
absorption
Pseudoephedrine GITS 180mg SD 12 healthy male High-fat breakfast After meal 96h plasma Rate and extent of CHAO et al. (1991)
and combination pseudoephedrine subjects absorption of both
brompheniramine tablet 10mg drugs unaffected
brompheniramine
Rifabutin Capsule 150mg SD 12 healthy male Standard breakfastb - 15 min after 168h Cmax reduced 16% NARANG et at.
subjects starting meal plasma,48h but other parameters (1992)
urine unchanged
Sparfioxacin 400mg SD 10 healthy Standard breakfast Concomitant 96h plasma Rate and extent of THEBAULT et at.
subjects intake absorption unaffected (1990)
Sparfioxacin 200mg SD 6 healthy 48h plasma Rate and extent of SAKASHITA et al.
subjects absorption unaffected (1991)
Temaftoxacin Crushed tablet 600mg SD 18 healthy male Osmolite enteral 200ml water Concomitant 48h plasma Rate and extent of LUBOWSKI et al.
via nasogastric subjects feeding solution with enteral absorption unaffected (1992)
tube feeding solution
Temafloxacin Tablet 600mg SD Healthy subjects Standard breakfastb 240ml water 20 min after 48h plasma T=, delayed from GRANNEMAN and
meal 1.8 to 3.0 h. Other MUKHERJEE (1992)
parameters unaffected
Theophylline Monospan 450mg SD 22 healthy male High-fat breakfastb 6 ounces Immediately 72h plasma Rate and extent of HARRISON et a1.
capsule subjects water after meal absorption unaffected (1993)
Theophylline Monospan 900mg 29 healthy male High-fat breakfast b 6 ounces Immediately 72h plasma Rate and extent of HARRISON et a1.
capsule subjects water after meal absorption unaffected (1993)
Theophylline Theo-24 6mg/kg SD 6 healthy male Ensure enteral 100ml water One hour after 72h plasma Rate and extent of PLEZIA et a1.
Capsules subjects feeding, as ten third enteral absorption unaffected (1990)
boluses feeding bolus by enteral feeding
Theophylline Uniphyl Variable, in 200-mg MD 22 patients Evening meal 45 min after Peak and Marginal increases in ARKINSTALL et at.
tablets increments 1900h (14M, 8F) with meal trough plasma Cm" and C rni, (1991)
each day chronic airways levels, values. No
obstruction together with differences in
pulmonary pulmonary function
function tests test results
Tiaprofenic acid Sustained 600mg SD 7 healthy Light breakfast or 100ml water After meal 24 h plasma Rate and extent of WILDING et al.
release tablet subjects, 5M, 2F heavy breakfast b absorption unaffected (1992)
with 152Sm by meals despite
slower gastric
emptying
Trimetazidine Sustained 80mg SD 12 healthy male Standard breakfast With meal 48h plasma Rate and extent of CHAUFOUR et al.
release subjects absorption unaffected (1992)
formulation
Verapamil Sustained 240mg SD 12 healthy male Standard breakfastb - 10min after 36h plasma No significant DEVANE et al.
release subjects meal differences in plasma (1990)
formulation profiles of verapamil DEVANE and
or nor-verapamil KELLY (1991)
Verapamil Pellet filled 240mg SD 32 healthy male Apple sauce 150ml water With meal 48h plasma No significant KOZLOSKI et al.
capsule, or volunteers differences in plasma (1992a)
pellets profiles of verapamil
sprinkled on or nor-verapail
food

a Single dose.
b Meal described.
C Multiple dose.
104 P.G. WELLING

::::;-
4
~ A
s<II
.is.. 3
c
:cQ
E 2
-<<U
E
c::'"
<U

20 40 60 80 100 120 140 160 180 200

TIME (hr)

10
::::;-
~
E
8
B
S
<II
c
.is.. 6
:cQ
E
-<<U 4
E
<II
<U
c:: 2

20 40 60 80 100 120 140 160 180 200

TIME (hr)

Fig. 20A,B. Mean plasma versus time concentrations of amlodipine in 12 subjects

following A a single 10-mg oral dose of amlodipine fasting (0) and following a standard
breakfast (e) and B a single 20-mg oral dose of amlodipine as a solution (0) and
capsule (L.) dose fasting. (From FAULKNER et al. 1989)

metoclopramide had a significant effect on plasma profiles of the ergot deriva-

tive bromocriptine (KOPITAR et al. 1991). Consistent with its effects on gastric
emptying, food caused a slight delay in bromocriptine absorption, while
metoclopramide had the opposite effect. However, both changes were trivial
and overall bioavailability was unaffected by either treatment.
While previously cited studies showed increases in absorption of the
cephalosporin prodrug cefetamet pivoxil from tablets (BLOUIN and STOECKEL
1993; TAM et al. 1990), a study using an oral syrup formulation of cefetamet
pivoxil showed no food effect when a 385-mg (cefetamet free acid equivalents)
dose of cefatemet pivoxil was administered together with a standard breakfast
(DUCHARME et al. 1993). Mean Cmax values in plasma were 2.7,ug/ml with food
compared to 2.9,ug/ml fasting and absolute bioavailability was 38% and 34%
after fed and fasting doses, respectively. Interestingly, the syrup had signifi-
cantly lower absolute systemic bioavailability than that of a tablet adminis-
Drug-Food Interactions Affecting Drug Absorption 105

tered under fed conditions. This is attributed to the lack of effect by food on
the bioavailability of cefetamet pivoxil from the syrup formulation. While the
mechanism for the different responses of cefetamet pivoxil tablets and syrup
to food is unknown, it is claimed that plasma profiles of unbound cefetamet are
adequate to treat susceptible organisms (DUCHARME et al. 1993). Absorption
of both cimetidine and ranitidine was unaffected when they were administered
to healthy male subjects after a standard breakfast compared to fasting
(DESMOND et al. 1990). In the fasted state, absorption of cimetidine was de-
creased 24%, and ranitidine 59%, when these compounds were taken together
with an antacid. In the fed state, however, co administered antacid did not have
the same effect on cimetidine or ranitidine absorption. It is proposed that the
antacid effect seen in the fasting state is related to impaired tablet dissolution
and to drug binding to unabsorbed antacid. Abolition of this effect by food
may be due to competition for drug-binding sites on the antacid. However, this
binding must be weak because of the lack of effect by food on drug absorption
in the absence of antacid.
Data on the effect of food on cyclosporine absorption are conflicting. In a
study cited in Table 5, bioavailability from a chocolate emulsion was 23% and
42% after low-fat and high-fat meals, respectively (GUPTA et al. 1990). Other
studies have reported no significant change (LINDHOLM et al. 1990) or de-
creased cyclosporine absorption with food (KEOWN et al. 1982). Significant
increases in cyclosporine absorption with food have been reported in renal
transplant patients (PTACHCINSKI et al. 1985). Two further studies have re-
ported little effect by food on cyclosporine absorption (HONCHARIK et al.
1991). In the first study, cyclosporine was administered to 14 renal transplant
patients immediately following a moderate- or trace-fat breakfast. Neither
meal had any significant effect on cyclosporine pharmacokinetic parameters.
Mean Cmax values were 410, 346, and 365ng/ml, and mean AVe values were
2115, 2085, and 2145 ng·h/ml following fasting, moderate-fat, and high-fat
treatments at an average cyclosporine dose of 3.51 mg/kg per day. In the
second study, conducted in healthy male subjects, a standard light breakfast
caused a 17% reduction in mean peak plasma cyclosporine levels, but had no
effect on area under plasma profiles, with mean values of 7283 and 7453ng·h/
ml from a 6-mg/kg dose to fasted and fed subjects, respectively (LINDHOLM et
al. 1990). Addition of bile salts to the nonfasted treatment, on the other hand,
caused mean Cmax values to increase 1.1-fold, and overall bioavailability
1.2-fold, relative to fasting. These results suggest that, in this study at least,
bile acid formation is an important determinant of cyclosporine absorption.
This observation does not explain the various outcomes of cyclosporine-food
interactions obtained in other studies.
In an attempt to further elucidate the mechanism of drug-food interac-
tions, and to identify an appropriate surrogate test meal, the absorption of
some widely divergent compounds, diazepam, propranolol, norethindrone,
and ethinyl estradiol, was examined in the presence of water, triglyceride oil,
and a sucrose polyester, Olestra (ROBERTS and LEFF 1989). The study was
106 P.G. WELLING

conducted in healthy subjects. The absorption of both diazepam (GREENBLATT

et al. 1978) and propranolol (MELANDER et al. 1977) has been previously
shown to be increased by solid meals, although a more recent study with
diazepam showed that neither the rate nor extent of absorption was altered by
food, either from a conventional formulation or a slow-release formulation
(Du SOVICH et al. 1990). The liquid preparations had no significant effect on
the absorption of any of the four compound studied, except that the time of
peak plasma diazepam levels was increased by triglyceride oil, relative to the
Olestra and water treatments. The inconsistency between results obtained
from the two earlier diazepam studies makes it difficult to assess the overall
predictability of the liquid formulation for actual diazepam-food interactions
in a clinical environment.
Food was shown to have no effect on the rate and extent of absorption of
the new cholinesterase inhibitor E2020 (MIHARA et al. 1993). Following single
2-mg p.o. doses of E2020 to healthy male volunteers, mean peak plasma E2020
levels of3.3 and 3.2ng/ml and AVe (0-168 h) values of166.5 and 172.8ng·h/ml
were obtained under fasting and fed conditions. The nature of the meal in this
case was not described but was presumably a standard Japanese-style break-
fast. Similar lack of effects are reported for the selective serotonin reuptake
inhibitor fluvoxamine (VAN HARTEN et al. 1991) and also the systemic avail-
ability of the S(+) and R( -) enantiomers of ibuprofen (LEVINE et al. 1992). On
the other hand, sucralfate, often administered to reduce gastrointestinal side
effects of nonsteroidal anti-inflammatory agents, reduced peak concentrations
of both enantiomers of ibuprofen. Neither food nor sucralfate had any influ-
ence on R( -) to S(+) enantiomer inversion. A report on the lack of food effect
on levodopa absorption illustrates, again, the complexity and unpredictability
of drug-food interactions. An earlier report described inhibition of levodopa
absorption, based on clinical results, by high-protein diets (GILLESPIE et al.
1973). Later studies, included in this review, described reduced levodopa
absorption after a standard luncheon (Roos et al. 1993) and increased absorp-
tion from both immediate and controlled release formulations following
a light breakfast (WILDING et al. 1991b). In a further study conducted in
healthy volunteers, a meal containing 30.5 g protein had no effect on levodopa
absorption from a 150-mg solution dose, whereas absorption was reduced by
10%, and peak plasma levodopa levels by 26%, when drug was ingested
following a meal containing only 10.5 g protein (ROBERTSON et al. 1991). The
poor bioavailability of levodopa following the low-protein meal, relative to the
fasting state, suggests that low-protein diets do not increase levodopa absorp-
tion. It is speculated that any beneficial effects of a low-protein diet on
levodopa efficacy for treatment of Parkinson's disease may be related to
reduced competition for transport across the blood-brain barrier rather than
increased systemic availability.
Absorption of the beta-adrenoreceptor antagonist metoprolol has previ-
ously been shown to be significantly increased by food (PHELAN et al. 1991).
The effect of food was further examined on a new controlled release formula-
Drug-Food Interactions Affecting Drug Absorption 107

tion of metoprolol succinate. Following a standard high-fat breakfast, peak

plasma levels and overall bioavailability of metoprolol from a single 400-mg
dose increased by only a small extent (SANDBERG et al. 1988). Following a
standard high-carbohydrate breakfast, plasma metoprolollevels from a 50-mg
dose were unaffected. Thus, it appears that absorption of metoprolol from this
controlled release formulation is affected far less by food than the conven-
tional release formulation studied previously.
Previous studies using single doses of controlled release formulations of
morphine have reported increased and also unchanged absorption in the
presence of food (KAIKO et al. 1990; HUNT and KAIKO 1991). In order to
determine the effect of food on controlled release morphine under clinical
dosing conditions, morphine sulfate was administered in sustained release
tablets at a dose level of 30mg every 12h for seven doses either immediately
after or 2h before standard caffeine-free meals (BASS et al. 1992). Plasma
morphine levels monitored during the repeated dosing and profiled during the
12h following the last dose were similar from fasted and fed treatments.
Profiles obtained during the period following the last morphine sulfate dose
are shown in Fig. 21. The AUe for fed subjects during the 12-h period was
9.8% less than that for fasted subjects. The lack of food effect observed in this
repeated dose study was observed using the same sustained release formula-
tion to that associated with increased absorption in the presence of food
following single doses (HUNT and KAIKO 1991).
Two separate studies showed that food does not affect the rate or extent
of absorption of the new imidazoline antihypertensive agent moxonidine.

30

::::i'
E
~
E 20
Q.I

:cQ.
C

0
-:;
to 10
E
'"
to
c;:

72 76 80 84
TIME (hr)

Fig. 21. Mean plasma versus time concentrations of morphine during the 72- to 84-h
dosing interval in 24 subjects following repeated 30-mg q 12h oral doses of morphine
sulfate as sustained release tablets administered under fasting (0) and fed (e) condi-
tions. (From BASS et al. 1992)
 108 P.G. WELLING

Following single 0.2-mg doses of moxonidine tablets to healthy male subjects,

mean peak plasma levels of moxonidine decrease 14% while systemic avail-
ability and urine excretion were essentially unchanged by food (THEODOR et al.
1992). In a second study in healthy male subjects, almost identical results
were obtained, with modest, nonsignificant, increases in peak moxonidine
plasma levels and no effect on absorption in the presence of food (WEIMANN
and RUDOLPH 1992). Similar to other quinolone antibiotics, absorption of
sparftoxacin appears to be essentially unaffected by food. Following single
400-mg doses to healthy subjects, the time of peak sparftoxacin plasma levels
was increased from 3.1 to 4.7h by food, but peak levels and systemic availabil-
ity were unaffected (THEBAULT et al. 1990). In a similar study conducted in
healthy individuals, plasma sparftoxcin levels following a single 200-mg dose
under fed and fasted conditions were almost superimposable (SAKASHITA et al.
1991). Similarly, the rate and extent of temaftoxacin absorption were essen-
tially unchanged when administered by nasogastric tube with or without
enteral feeding (LUBOWSKI et al. 1992), and from conventional tablets admin-
istered as a single 200-mg dose under fasting and fed conditions (GRANNEMAN
and MUKHERJEE 1992). Plasma profiles of temaftoxacin from 2 x 200-mg tablets
and 1 x 400-mg tablet fasting, and 1 x 400-mg tablet nonfasting, are shown in
Fig. 22.
Incorporating the nonsteroidal-anti-inftammatory agent tiaprofenic acid
into a novel sustained release system comprising discrete sustained release

3.0

2.5

I
~

2.0
c
·u
'><"
0
0;: 1.5
'"
E
<II
I-

'"
E
<I>
1.0
'"
~

0.5

0.0
0 12 24 36 48
Time (hr)

Fig. 22. Mean plasma versus time concentrations of temafloxacin in 18 subjects follow-
ing a single oral dose of 1 x 400-mg tablet fasting (0),2 x 200-mg tablets fasting (0),
and 1 x 400-mg tablet after food (e). (From GRANNEMAN and MUKHERJEE 1992)
Drug-Food Interactions Affecting Drug Absorption 109

pellets in a single tablet gave rise to absorption profiles that were unaffected by
food (WILDING et al. 1992). Healthy volunteers received 600mg drug in the
fasting state and also after standard light and heavy breakfasts. Gamma
scintigraphy measurements showed that stomach emptying of 152Sm-I abeled
pellets was delayed by both light and heavy meals. Despite the delayed stom-
ach emptying, plasma profiles of tiaprofenic acid were almost superimposable
after fasting and fed treatments.
Although food has recently been shown to delay (KANN et al. 1989; YUEN
et al. 1993) and also to increase (COOK et al. 1990) absorption of theophylline
from controlled release formulations, other studies have shown little effect.
The rate and extend of theophylline absorption from Monospan capsules, in
which drug is contained in I-mm-diameter beadlets, were almost identical
when administered to healthy volunteers fasting or immediately after a high-
fat breakfast (HARRISON et al. 1993). Similarly, the systemic availability of
theophylline from Theo-24 capsules was unaffected by enteral liquid feeding
in healthy male subjects (PLEZIA et al. 1990). Earlier reports on the lack of
effect by food on theophylline absorption from Uniphyl tablets during chronic
evening dosing were confirmed in a subsequent study which monitored both
the serum drug levels and also clinical efficacy in patients with chronic airways
obstruction (ARKINSTALL et al. 1991). Following mean daily (evening) theo-
phylline doses of 818mg, peak plasma theophylline concentrations were
14.4.ug/ml with food and 13.1.ug/ml fasting. Mean trough levels were 7.4 and
6.9.ug/ml. There were no treatment-related differences in pulmonary function
tests including spirometry, asthma symptom scores, side effects, or the need
for beta-agonist inhalers.
The last compound to be discussed in this section is of interest largely
because of the divergent results obtained regarding food interactions with
controlled release preparations. Verapamil absorption was previously shown
to be reduced by 30% and peak serum levels by 48% when a sustained release
tablet was taken with food (HOON et al. 1992). However, different results were
obtained when verapamil was administered in a different sustained release
formulation. Twelve healthy male volunteers received single 200-mg ve-
rapamil doses in a newly marketed sustained release formulation employing
rate-controlling technology in the form of a capsule, either fasting or 10min
after a standard breakfast (DEVANE et al. 1990; DEVANE and KELLY 1991).
Plasma profiles from the fed and fasting treatments, together with verapamil
profiles from a conventional 80-mg dose administered every 8h are shown in
Fig. 23. Plasma profiles of verapamil following the sustained release capsule
were superimposable in the fed and fasting states. Plasma profiles of the
metabolite norverapamil were similarly unaffected by food. Mean peak
plasma verapamil plasma levels were 77 ng/ml following both sustained release
treatments, compared to 95 ng/ml following the conventional tablet. Areas
under plasma verapamil concentration versus time curves from zero to infinite
time were 1541 and 1387ng·h/ml from fasted and fed sustained release doses,
respectively, compared to 1370ng·h!ml from the conventional tablet. The
110 P.G. WELLING

80

70

::; 60

~
S 50
·e
tel
c.. 40
~
III
>
tel 30
E
'"
tel
li: 20

10

12 24 36 48
Time (hr)

Fig.23. Mean plasma versus time concentrations of verapamil in 12 subjects following

once daily oral doses of a 240-mg sustained release verapamil capsule under fasting (0)
and fed (e) conditions and after three times daily oral doses of an 80-mg immediate
release tablet fasting (0). (From DEVANE and KELLY 1991)

divergent results obtained from these studies are most likely related to the
formulations used. Release of drug from membrane controlled release formu-
lations, including those using osmotic pump technology, are generally less
sensitive to the gastrointestinal environment, including the presence of food,
than other types of controlled release formulations. The advantage of a rate-
controlling membrane dosage form for verapamil was further demonstrated in
a study in which drug was administered as extended release pellets in the
fasting state and also sprinkled on apple sauce. There were no significant
differences in plasma verapamil or norverapamil profiles between the fasting
and apple sauce treatments (KOZLOSKI et al. 1992a).

I. Conclusions
The number of articles published during the 5 years covered by this review
illustrates the high level of interest, from scientific, regulatory, and clinical
perspectives, on interactions between drugs and ingested food influencing
drug absorption. As in the initial review on this topic (WELLING 1977), drugs
and drug formulations continue to fall naturally into four major categories of
those whose absorption is reduced, delayed, increased, or not affected by food.
In this review, a fifth category of drugs whose absorption is accelerated but not
increased by food was identified.
Grouping of drugs and formulations in this way is entirely empirical. The
interactions could have been grouped differently, for example, among thera-
Drug-Food Interactions Affecting Drug Absorption 111

peutic classes or according to physicochemical characteristics. Whichever way

the interactions are sorted, one is nonetheless always confronted with the
tremendous divergence of results that occur, largely independent of therapeu-
tic class and of physicochemical properties. This is the conundrum that pre-
sents to those who concern themselves with drug absorption, circulating drug
profiles, and, inevitably, clinical efficacy consequences. Among the reports
summarized in this review, approximately 57 reported delayed drug absorp-
tion as a result of food interactions, 40 reported reduced absorption, 49
reported increased absorption, 3 reported accelerated absorption, and 46
reported no interaction. Many of the drugs whose absorptin was reduced or
increased also exhibited delayed absorption to varying extents.
The last decade may be considered to have been a data-collecting period.
Altogether there have been approximately 400 studies reported in the litera-
ture describing drug-food interactions, and the patterns of interaction have not
changed significantly with time. No new mechanisms or pivotal relationships
have been described, and one can only conclude that, of all the possible
interactions that may occur and which are summarized in Tables 1 and 2, any
interaction that predominates will determine the final observed effect. Results
of various reported studies have indicated that the nature of a drug-food
interaction may be at least partially predictable from a physicochemical per-
spective, but accurate predictability is plagued by many exceptions, which are
often spectacular in nature. The lack of predictability of the outcome of drug-
food interactions is such that this phenomenon must always be considered, and
is in fact becoming mandatory in all drug development programs, and in
clinical use.
The clinical impact of a drug-food interaction, as stated earlier, depends
on the nature and degree of change in circulating drug levels, the margin of
safety, and the slope of the drug concentration-response curve. A small change
in circulating drug levels for a drug with a relatively flat dose response curve is
likely to be of little clinical consequence. On the other hand, a large change in
circulating drug levels for a drug with a relatively steep dose response curve, or
a relative narrow safety margin, may be critical.
While it continues to be necessary to explore drug-food interactions for
drugs and drug formulations, additional avenues need to be explored to seek
mechanistic patterns that may lead to better prediction of the nature and
extent of changes in circulating drug levels due to the presence of food, and
their possible clinical impact.

References
Albertioni F, Juliusson G, Liliemark J (1993) On the bioavailability of 2-chloro-2'-
deoxyadenosine (CdA); the influence of food and omeprazole. Eur J Clin
PharmacoI44:579-582
Arkinstall WW, Brar PS, Stewart JH (1991) Repeated dosing of Uniphyl tablets under
fed and fasting conditions: comparison of serum theophylline levels, pulmonary
function and asthma symptoms. Ann Allergy 67:583-587
112 P.G. WELLING

Astarloa R, Mena MA, Sanchez V, De La Vega L, De Yebenes JG (1992) Clinical and

pharmacokinetic effects of a diet rich in insoluble fiber on Parkinson's disease.
Clin NeuropharmacoI15:375-380
Axelson JE, Chan GLY, Kirsten EB, Mason WD, Lanman EC, Kerr CE (1987)
Enhancement of the bioavailability of propafenone by food. Br J Clin Pharmacol
23:735-741
Bailey DG, Spence JD, Munoz C, Arnold JMO (1991) Interaction of citrus juices with
felodipine and nifedipine. Lancet 337:268-269
Bannwarth B, Combes C, Vin~on G, Bouchet JL, Aparicio M, Begaud B (1992)
Influence of low protein diet on the pharmacokinetics of carbamazepine in chronic
renal failure. Fundam Clin Pharmacol 6:222
Barbhaiya RH, Patel RB, Corrick-West HP, Joslin RS, Welling PG (1982) Compara-
tive bioavailability of hydrochlorothiazide from oral tablet dosage forms, deter-
mined by plasma and urinary excretion models. Biopharm Drug Dispos 3:329-336
Barbhaiya RH, Shukla UA, Gleason CR, Shyu WC, Pittman KA (1990) Comparison of
the effects of food on the pharmacokinetics of cefprozil and cefaclor. Antimicrob
Agents Chemother 34:1210-1213
Barnwell SG, Laudanski T, Dwyer M, Story MJ, Guard P, Cole S, Attwood D (1993)
Reduced bioavailability of atenolol in man: the role of bile acids. Int J Pharm
89:245-250
Barone JA, Koh JG, Bierman RH, Colaizzi JL, Swanson KA, Gaffar MC, Moskovitz
BL, Mechlinski W, Van de Velde V (1993) Food interactions and steady state
pharmacokinetics of itraconazole capsules in healthy male volunteers. Antimicrob
Agents Chemother 37:778-784
Bass J, Shepard KV, Lee JW, Hulse J (1992) An evaluation of the effect of food on the
oral bioavailability of sustained-release morphine sulfate tablets (Oramorph SR)
after mUltiple doses. J Clin PharmacoI32:1003-1007
Basu TK (1988) Drug-nutrient interactions. Croom Helm, London
Beckermann B, Beneke M, Bottcher M, Dietrich H, Horstmann R, Seitz I (1993)
Influence of formulation, food or antacids on the pharmacokinetics of BAY X
1005 in human volunteers. Arch Pharmacol 347 [Suppl]:R27
Beerman B, Groschinsky-Grind M (1978) Enhancement of the gastrointestinal absorp-
tion of hydrochlorothiazide by propantheline. Eur J Clin Pharmacol 13:385-387
Belch JJF, McLaren M, Lau CS, MacKay IR, Bancroft A, McEwen J, Thompson JM
(1993) Cicaprost, an orally active prostacyclin analogue: its effects on platelet
aggregation and skin blood flow in normal volunteers. Br J Clin PharmacoI35:643-
647
Bergmann JH, Simonneau G, Chassany 0, Caulin C (1992) Effect de la prise d'un repas
et d'un anti acide sur la cinetique de l'acetorphan. Therapie 47:229
Bieck PR, Antonin K-H, Schmidt E (1993) Clinical pharmacology of reversible
monoamine oxidase-A inhibitors. Clin Neuropharmacol16 [SuppI2]:S34-S41
Blouin RA, Stoeckel K (1993) Cefetamet pivoxil clinical pharmacokinetics. Drug
Dispos 25:172-188
Brookes LG, Sambol NC, Lin ET, Gee W, Benet LZ (1991) Effect of dosage form,
dose and food on the pharmacokinetics of metformin. Pharm Res 8 [Suppl]:S320
Chao ST, Prather D, Pinson D, Coen P, Pruitt B, Knowles M, Place V (1991) Effect of
food on bioavailability of pseudoephedrine and brompheniramine administered
from a gastrointestinal therapeutic system. J Pharm Sci 80:432-435
Charman WN, Rogge MC, Boddy A W, Berger BM (1993) Effect of food and a
monoglyceride emulsion formulation on danazol bioavailability. J Clin Pharmacol
33:381-386
Chaufour S, Funck-Brentano C, Jaillon P (1992) Pharmacokinetics of trimetazidine SR
in healthy volunteers: influence of food on oral drug disposition. Fundam Clin
Pharmacol 6:225
Choi RL, Sun JX, Kochak GM (1993) The effect of food on the relative bioavailability
of fadrazole hydrochloride. Biopharm Drug Dispos 14:779-784
Drug-Food Interactions Affecting Drug Absorption 113

Chu S-Y, Park Y, Locke C, Wilson DS, Cavanaugh JC (1992) Drug-food interaction
potential of clarithromycin, a new macrolide antimicrobial. J Clin Pharmacol
32:32-36
Clements JA, Heading RC, Nimmo WS, Prescott CF (1978) Kinetics of acetaminophen
absorption and gastric emptying in man. Clin Pharmacol Ther 24:420-431
Cole AFD, Baxter JG, Jackson BJ, Hew-Wing P, Guntert TW, Lalka D (1992) Phar-
macokinetic and metabolic aspects of the moclobemide-food interaction. Psycho-
pharmacology 106:S37-S39
Cole SK, Story MJ, Laudanski T, Dwyer M, Attwood D, Robertson J, Barnwell SG
(1992) Targeting drugs to the enterohepatic circulation: a potential drug delivery
system designed to enhance the bioavailability of indomethacin. Int J Pharm
80:63-73
Confalonieri S, Cosmi G, Guiso G, Gherardi S, Guido M, Caccia S (1992) The pharma-
cokinetics of a potential memory-enhancing compound, CL 275838, in fasted and
fed volunteers. Drug Invest 4:322-328
Conway EL, McNeil JJ, Hurley J, Jackman GP, Krum H, Howes LG, Louis WJ (1993)
The effects of food on the oral bioavailability of doxazosin in hypertensive sub-
jects. Drug Invest 6:90-95
Cook CS, Hauswald CL, Grahn A Y, Kowalski K, Karim A, Koch R, Schoenhard GL,
Oppermann JA (1990) Suitability of the dog as an animal model for evaluating
theophylline absorption and food effects from different formulations. Int J Pharm
60:125-132
Cook HJ, Mundo CR, Fonseca L, Gasque L, Moreno-Esparza R (1993) Influence of
the diet on bioavailability of tetracycline. Biopharm Drug Dispos 14:549-553
De Mey C, Meineke I (1992) Prandial and diurnal effects on the absorption of orally
administered enteric coated 5-aminosalicylic acid (5-ASA). Br J Clin Pharmacol
33:179-182
Degen PH, Cardot JM, Czendlik C, Dieterle W (1993) Influence of food on the
disposition of the monoamine oxidase-A inhibitor brofaromine in healthy volun-
teers. Biopharm Drug Dispos 14:209-215
Degen PH, Flesch G, Cardot J-M, Czendlik C, Dieterle W (1994) The influence of food
on the disposition of the anti epileptic oxcarbazepine and its major metabolites in
healthy volunteers. Biopharm Drug Dispos 15:519-526
Desante KA, Zeckel ML (1992) Pharmacokinetic profile of loracarbef. Am J Med
92:16S-19S
Desmond PV, Harman PJ, Gannoulis N, Kamm M, Mashford ML (1990) The effect of
an antacid and food on the absorption of cimetidine and ranitidine. J Pharm
Pharmacol 42:352-354
Devane JG, Kelly JG (1991) Effect of food on the bioavailability of a multiparticulate
sustained-release verapamil formulation. Adv Ther 8:48-53
Devane JG, Kelly JG, Geoghegan B (1990) Pharmacokinetic and in vitro
characteristics of sustained release verapamil products. Drug Dev Ind Pharm
16:1233-1248
DeVries TM, Voigtman RE, Posvar EL, Nesbitt RU, Forgue ST (1991) Pharmacoki-
netic characterization of a procainamide tablet suitable for twice daily dosing.
Pharm Res 8 [Suppl]:S315
Doose DR, Gisclon LG, Stellar SM, Riffits JM, Hills JF (1992a) The effect of food on
the bioavailability of topiramate from 100- and 400-mg tablets in healthy male
subjects. Epilepsia 33 [Suppl 3]:105
Doose DR, Minn FL, Stellar S, Nayak RK (1992b) Effects of meals and meal compo-
sition on the bioavailability of fenretidine. J Clin Pharmaco132:1089-1095
Dressman JB, Berardi RR, Elta GH, Gray TM, Montgomery P A, Lau HS, Pelekoudas
KL, Szpunar GL, Wagner JG (1992) Absorption of flurbiprofen in the fed and
fasted states. Pharm Res 9:901-907
Drew RH, Gallis HA (1992) Azithromycin-spectrum of activity, pharmacokinetics, and
clinical applications. Pharmacotherapy 12:161-173
114 P.G. WELLING

Du Souich P, Lery N, Lery L, Varin F, Boucher S, Vezina M, Pilon D, Spenard J, Caille

G (1990) Influence of food on the bioavailability of diltiazem and two of its
metabolites following the administration of conventional tablets and slow-release
capsules. Biopharm Drug Dispos 11:137-147
Ducharme MP, Edwards DJ, McNamara PJ, Stoeckel K (1993) Bioavailability of syrup
and tablet formulations of cefetamet pivoxil. Antimicrob Agents Chemother
37:2706-2709
Dudley MN, Marchbanks CR, Flor SC, Beals B (1991) The effect of food or milk on the
absorption kinetics of ofloxacin. Eur J Clin PharmacoI41:569-571
Dupuis LL, Koren G, Silverman ED, Laxer RM (1992) Influence of food on the
bioavailability of oral methotrexate (MTX). Arthritis Rheum 35 [Suppl]:S57
Edgar B, Bailey D, Bergstrand R, Johnsson G, Regardh CG (1992) Acute effects of
drinking grapefruit juice on the pharmacokinetics and dynamics of felodopine and
its potential clinical relevance. Eur J Clin PharmacoI42:313-317
Ehrsson H, Wallin I, Ros AM, Eksborg S, Berg M (1994) Food-induced increase in
bioavailability of 5-methoxypsoralen. Eur J Clin Pharmacol 46:375-377
Eller MG, Della-Coletta AA (1990) Absence of effect of food on alprazolam absorp-
tion from sustained release tablets. Biopharm Drug Dispos 11:31-37
Eller MG, Walker BJ, Yuh L, Antony KK, McNutt BE, Okerholm RA (1992) Absence
of food effects on the pharmacokinetics of terfenadine. Biopharm Drug Dispos
13:171-177
Ewe K, Press AG, Dederer W (1989) Gastrointestinal transit of undigestible solids
measured by metal detector EAS II. Eur J Clin Invest 19:291-297
Faulkner JK, Hayden ML, Chasseaud LF, Taylor T (1989) Absorption of amlodipine
unaffected by food. Arzneimittelforschung 39:799-801
Fiese EF, Steffen SH (1990) Comparison of the acid stability of azithromycin and
erythromycin A. J Antimicrob Chemother 25 [Suppl A]:39-47
Fowles SE, Fairless AJ, Pierce DM, Prince WT (1991) A further study of the effect of
food on the bioavailability and pharmacokinetics of penciclovir after oral admin-
istration of famciclovir. Br J Clin Pharmacol 30:657P-658P
Fowles SE, Pierce DM, Prince WT, Thow JC (1990b) Effect of food on the
bioavailability and pharmacokinetics of penciclovir, a novel antiherpes agent,
following oral administration of the prodrug, famciclovir. Br J Clin Pharmacol
29:620P-621P
Frishman WH (1993) A new extended-release formulation of diltiazem HCI for the
treatment of mild-to-moderate hypertension. J Clin Pharmacol 33:612-622
Frost RW, Carlson JD, Dietz AJK, Heyd A, Lettieri JT (1989) Ciprofloxacin pharma-
cokinetics after a standard or high-fat/high-calcium breakfast. J Clin Pharmacol
29:953-955
Frydman A (1992) Pharmacokinetic profile of nicorandil in humans: an overview.
J Cardiovasc Pharmacol20 [SuppI3]:S34--S44
Fujimaki Y, Sudo K, Hakusui H, Tachizawa H, Murasaki M (1992) Single- and mul-
tiple-dose pharmacokinetics of nefiracetam, a new nootropic agent, in healthy
volunteers. J Pharm PharmacoI44:750-754
Gan V, Chu S-Y, Kusmiesz HT, Craft JC (1992) Pharmacokinetics of a clarithromycin
suspension in infants and children. Antimicrob Agents Chemother 36:2478-2480
Gertz BJ, Holland SD, Kline WF, Matuszewski BK, Porras AG (1993) Clinical phar-
macology of alendronate sodium. Osteoporosis Int [Suppl 3]:S13-S16
Gillespie NG, Mena I, Cotzias GC, Bell MA (1973) Diets affecting treatment of
parkinsonism with levodopa. J Am Diet Assoc 62:525-528
Granneman GR, Mukherjee D (1992) The effect of food on the bioavailability of
temafloxacin. A review of 3 studies. Clin Pharmacokinet 22 [Suppl1]:48-56
Greb WH, Brett MA, Buscher G, Dierdorf H-D, von Schrader HW, Wolf D, Mellows
G, Zussman BD (1989) Absorption of paroxetine under various dietary conditions
and following antacid intake. Acta Psychiatr Scand 80 [SuppI350]:99-101
Greenblatt DJ, Allen MD, MacLaughlin DS, Harmatz JS, Shader RI (1978) Diazepam
absorption: effect of antacids and food. Clin Pharmacol Ther 24:600-609
Drug-Food Interactions Affecting Drug Absorption 115

Gupta SK, Benet LZ (1990) High-fat meals increase the clearance of cyclosporine.
Pharm Res 7:46-48
Gupta SK, Manfro RC, Tomlanovich SJ, Gambertoglio JG, Garovoy MR, Benet LZ
(1990) Effect of food on the pharmacokinetics of cyclosporine in healthy subjects
following oral and intravenous administration. J Clin Pharmacol 30:643-653
Haegele KD, Hinze C, Jode-Ohlenbusch A-M, Creamer G, Borlack J (1991) Effects of
a standardized meal on the pharmacokinetics of the new cardiotonic agent
piroimone. Arzneimittelforschung 41:1225-1229
Hamaguchi T, Shinkuma D, Irie T, Yamanaka Y, Morita Y, Iwamoto B, Miyoshi K,
Mizuno N (1993) Effect of a high-fat meal on the bioavailability of phenytoin in a
commercial powder with a large particle size. Int J Clin Pharmacol 31:326-330
Hanyok 11 (1993) Clinical pharmacokinetics of sotalol. Am J CardioI12:19A-26A
Harrison LI, Riedel DJ, Armstrong KE, Goldlust MB, Ekholm BP (1992) Effect of
food on salsalate absorption. Ther Drug Monit 14:87-91
Harrison LI, Mitra AK, Kehe CR, Klinger NM, Wick KA, McCarville SE, Cooper KM,
Chang SF, Roddy PJ, Berge SM, Kisicki JC, Dockhorn R (1993) Kinetics of
absorption of a new once-a-day formulation of theophylline in the presence and
absence of food. J Pharm Sci 82:644-648
Hilleman DE, Mohiuddin SM, Destache CJ, Stoysich AM, Nipper HC, Maleskar MA
(1992) Impact of food on the bioavailability of encainide. J Clin PharmacoI32:833-
837
Hoke JF, Chi EM, Anthony KK, Kulmala HK, Sussman NM, Okerholm RA (1991)
Effect of food on the bioavailability of vigabatrin tablets. Epilepsia 32 [SuppI3]:6-
7
Holmes B, Brogden RN, Richards DM (1985) Norfioxacin: a review of its antimicrobial
activity, pharmacokinetic properties and therapeutic use. Drugs 30:482-513
Honcharik N, Yatscoff RW, Jeffery JR, Rush DN (1991) The effect of meal composi-
tion on cyclosporine absorption. Transplantation 52:1087-1089
Honma A, Ikeda K, Udo N, Sdamori M, Hasegawa K, Kuruma I, Uto M, Nakahara T,
Fujita T (1986) Aniracetam: clinical phase I study of aniracetam. J Clin Ther Med
2:929-952
Hoon TJ, McCollam PL, Beckman KJ, Hariman RJ, Bauman JL (1992) Impact of food
on the pharmacokinetics and electrocardiographic effects of sustained release
verapamil in normal sUbjects. Am J Cardiol 70:1072-1076
Hooper WD, Dickinson RG, Eadie MJ (1990) Effect of food on absorption of
lomefioxacin. Antimicrob Agents Chemother 34:1797-1799
Hopkins A (1966) The pattern of gastric emptying: a new review of old results. J
Physiol (Lond) 182:144-149
Hughes WT, Kennedy W, Shenep JL, Flynn PM, Hetherington SV, Fullen G,
Lancaster DJ, Stein DS, Palte S, Rosenbaum D, Lian SHT, Blum MR, Rugers M
(1991) Safety and pharmacokinetics of 566C80, a hydroxynaphthoquinone with
anti-Pneumocystis carinii activity: a phase I study in human immunodeficiency
virus (HIV)-infected man. J Infect Dis 163:843-848
Hunt TL, Kaiko RF (1991) Comparison of the pharmacokinetic profiles of two oral
controlled-release morphine formulations in healthy young adults. Clin Ther
13:482-488
Hussey EK, Donn KH, Powell JR, Lahey AP, Pakes GE (1991) Albuterol extended-
release products: effect of food on the pharmacokinetics of single oral doses of
Volmax® and Proventil® repetabs in healthy male volunteers. J Clin Pharmacol
31:561-564
Hyneck MA (1992) An overview of the clinical pharmacokinetics of nabumetone. J
Rheumatol 19 [Suppl 36]:20-24
Ingwersen SH, Mant TGK, Larson 11 (1993) Food intake increases the relative oral
bioavailability of vanoxerine. Br J Clin PharmacoI35:308-310
Iseki K, Satoh Y, Sugawara M, Miyazaki K (1994) Effect of Enterued® administration
on the intestinal absorption of orally active cefem antibiotics. J Pharm Soc Jpn
114:233-240
116 P.G. WELLING

Jallad NS, Callejas HJ, Woo-Ming RB, Weidler DJ (1991) The pharmacokinetics of
astemizole plus pseudoephedrine when taken with a standardized meal. J Clin
Pharmacol31:863
Jarvinen A, Nykanen S, Mattila J, Haataja H (1992) Effect of food on absorption and
hydrolysis of erythromycin acistrate. Arzneimittelforschung 42:73-76
Johnson BF, O'Grady J, Sabey GA, Bye C (1978) Effect of a standard breakfast on
digoxin absorption in normal subjects. Clin Pharmacol Ther 23:315-319
Kaiko R, Grandy R, Thomas G, Goldenheim P (1990) A single-dose study of the effect
of food ingestion and timing of dose administration on the pharmacokinetic profile
of 30mg sustained-release morphine sulfate tablets. Curr Ther Res 47:869-878
Kann J, Levitt MJ, Horodniak JW, Pay JW (1989) Food effects on the nighttime
pharmacokinetics of Theo-Dur tablets. Ann Allergy 63:282-286
Keown PA, Stiller CR, Laupacis AL, Howson W, Coles R, Stawecki M, Koegler J,
Carruthers G, McKenzie N, Sinclair NR (1982) The effects and side effects of
cyclosporine: relationships to drug pharmacokinetics. Transplant Proc 14:659-661
KivistO KT, Ojala-Karlsson P, Neuvonen PJ (1992) Inhibition of norfloxacin absorp-
tion by dairy products. Antimicrob Agents Chemother 36:489-491
Klamerus KJ, Troy SM, Ben-Maimon CS, Chiang ST (1993) Effect of age, gender, and
food on zalospirone disposition. Clin Pharmacol Ther 53:193
Kleinbloesem CH, Ouwerkerk M, Brodenfeldt R (1993) Food effect with extended
release formulations: nifedipine. Clin Pharmacol Ther 53:207
Knupp CA, Milbrath R, Barbhaiya RH (1993) Effect of time of food administration on
the bioavailability of didanosine from a chewable tablet formulation. J Clin
Pharmacol 33:568-573
Kopitar Z, Vrhovac B, Povsic L, Plavsic F, Francetic I, Urbancic J (1991) The effect of
food and metoclopramide on the pharmacokinetics and side effects of
bromocriptine. Eur J Drug Metab Pharmacokinet 16:177-181
Korth-Bradey JM, Fruncillo RJ, Klamerus K-J, Chiang ST, Conrad KA (1992) The
influence of high and low fat meals on single dose pharmacokinetics of
zalospirone. Pharm Res 9 [Suppl]:S326
Kosoglou T, Kazierad D, Schentag JJ, Patrick JE, Heimark L, Flannery BE, Affrime
MB (1993) Effect of food and gastric emptying rate (GER) on the bioavailability
(BA) of 5-ISMN. Clin Pharm Ther 53:206
Kozloski GD, de Vito JM, Johnson JB, Holmes GB, Adams MA, Hunt TL (1992a)
Bioequivalence of verapamil hydrochloride extended-release pellet-filled capsules
when opened and sprinkled on food and when swallowed intact. Clin Pharm
11:539-542
Kozloski GD, De Vito JM, Kisicki JC, Johnson JB (1992b) The effect of food on the
absorption of methotrexate sodium tablets in healthy volunteers. Arthritis Rheum
35:761-764
Kuhn M (1993) Drug interactions and their nursing implications. J NY State Nurses
Assoc 24:10-16
Lafontaine D, Mailhot C, Vermeulen M, Bissonnette B, Lambert C (1990) Influence of
chewable sucraflate or a standard meal on the bioavailability of naproxen. Clin
Pharm 9:773-777
Lasswell AB, Loreck ES (1992) Development of a program in accord with JACAHO
standards for counseling on potential drug-food interactions. J Am Diet Assoc
92: 1124-1125
Latini R, Tognoni G, Kates RE (1984) Clinical pharmacokinetics of amiodarone. Clin
Pharmacokinet 9:136-156
Lecaillon JB, Dubois JP, Soula G, Pichard E, Poltera AA, Ginger CD (1990) The
influence of food on the pharmacokinetics of CGP 6140 (amocarzine) after oral
administration of a 1200mg single dose to patients with onchocerciasis. Br J Clin
Pharmacol 30:629-633
Lecaillon JB, Poltera AA, Zea-Flores G, De Ramirez I, Nowell De Arevalo A (1991)
Influence of food related to dose on the pharmacokinetics of amocarzine and of its
Drug-Food Interactions Affecting Drug Absorption 117

N-oxide metabolite CGP 13231, after oral administration to 20 onchocerciasis

male patients from Guatemala. Trop Med Parasitol 42:286-290
Levine MAH, Walker SE, Paton TW (1992) The effect of food or sucralfate on the
bioavailability of S(+) and R(-) enantiomers of ibuprofen. J Clin Pharmacol
32:1110-1114
Liao S, Hills J, Stubbs RJ, Nayak RK (1992) The effect of food on the bioavailability
of tramadol. Pharm Res 9 [Suppl]:S308
Liedholm H, Wfthlin-Boll E, Melander A (1990) Mechanisms and variations in the
food effect on propranolol bioavailability. Eur J Clin Pharmacol 38:460-475
Lindholm A, Henricsson S, Dahlqvist R (1990) The effect of food and bile acid
administration on the relative bioavailability of cyclosporin. Br J Clin Pharmacol
29:541-548
Lode H, Stahlmann R, Koeppe P (1979) Comparative pharmacokinetics of cephalexin,
cefaclor, cefadroxil, and CGP 9000. Antimicrob Agents Chemother 16:1-6
Lohmann A, Dingler E, Sommer W, Schaffier K, Wober W, Schmidt W (1992)
Bioavailability of vinpocetine and interference of the time of application with food
intake. Arzneimittelforschung 42:914--917
Lotterer E, Ruhnke M, Trauemann M, Beyer R, Bauer FE (1991) Decreased and
variable systemic availability of zidovudine in patients with AIDS if administered
with a meal. Eur J Clin PharmacoI40:305-308
Lubowski TJ, Nightingale CH, Sweeney K, Quintiliani R (1992) The relative
bioavailability of temafloxacin administered through a nasogastric tube with and
without enteral feeding. Clin Pharmacokinet 22 [Suppl 1]:43-47
Lucas S, Rowinsky E, Wargin W, Hohneker J, Hsieh A, Donebower R (1993) Results
of a study of the effect of food on the bioavailability and pharmacokinetics of
Navelbine® liquid-filled soft gelatin capsules. Proc Am Soc Clin Pharmacol12:160
MacGowan AP, Greig MA, Andrews JM, Reeves DS, Wise R (1989) Pharmacokinetics
and tolerance of a new film-coated tablet of sodium fusidate administered as a
single oral dose to healthy volunteers. J Antimicrob Chemother 23:409-415
MacKay J, Mackie AE, Palmer JL, Moult A, Baber NS (1991) Investigation into the
mechanism for the improved oral systemic bioavailability of cefuroxime from
cefuroxime axetil when taken after food. Br J Clin Pharmacol 33:226P-227P
Margalith D, Duroux P, Bauerfeind P, Emde C, Koelz H-R, Biollaz J, Klotz U,
Armstrong D, Blum AL (1991) Famotidine should be taken with supper: the effect
of drug-meal interactions on gastric acidity and plasma famotidine levels. Eur J
Gastroenterol HepatoI3:405-412
Mattila MJ, Neuvonen PJ, Gothoni G, Hackman R (1972) Interference of iron pre-
parations and milk with the absorption of tetracyclines. In: Backer SB, Neuhaus
GA (eds) Toxicological problems of drug combinations. Excerpta Medica,
Amsterdam, pp 128-133
McLachlan AJ, Cutler DJ (1991) Bioavailability of hydroxychloroquine in rheumatoid
arthritis patients using deconvolution techniques. Aust J Hosp Pharm 21:333
McLean AJ, McNamara PJ, DuSouich P, Gibaldi M, Lalka D (1978) Food, splanchnic
blood flow and bioavailability of drugs subject to first-pass metabolism. Clin
Pharmacol Ther 24:5-10
McNeil JJ, Drummer OH, Raymond K, Conway EL, Louis WJ (1991) The influence of
food on the oral bioavailability of terazosin. Br J Clin PharmacoI32:775-776
Melander A, McLean A (1983) Influence of food intake on presystemic clearance of
drugs. Clin Pharmacokinet 8:286-296
Melander A, Danielson K, Schersten B, Wahlin E (1977) Enhancement of the
bioavailability of propranolol and metoprolol by food. Clin Pharmacol Ther
22:108-112
Melander A, Stenberg P, Liedholm H, Scherston B, Wahlin-Boll E (1979) Food-
induced reduction in bioavailability of atenolol. Eur J Clin PharmacoI16:327-330
Mengel H, Gustavson LE, Soerensen HJ, McKelvy JF, Pierce MW, Houston A (1991)
Effect of food on the bioavailability of tiagabine HCl. Epilepsia 32 [Suppl 3]:6
118 P.G. WELLING

Mihara M, Ohnishi A, Tomono Y, Hasegawa J, Shimamura Y, Yamazaki K, Morishita

N (1993) Pharmacokinetics of E2020 a new compound for Alzheimer's disease in
healthy male volunteers. Int J Clin Pharmacol 31:223-229
Monk JP, Campoli-Richards DM (1987) Ofloxacin: a review of its antibacterial activity,
pharmacokinetic properties and therapeutic use. Drugs 33:346-391
Motohiro T, Handa S, Yamada S, Oki S, Tsumura N, Yoshinaga Y, Sasaki H, Aramaki
M, Oda K, Kawakami A, Koga T, Sakata Y, Nishiyama T, Yamashita F, Hayashi
M, Amamoto M, Murakami T, Ono E, Shimizu T, Fukuda T, Watanabe J, Nagai
K, Sueyoshi K, Morita J, Hashimoto N, Yuge K, Kuda N, Ando H, Dai S,
Tominaga K (1992) Pharmacokinetics and clinical effects of cefdinir 10% fine
granules in pediatrics. Jpn J Antibiot 45:74-86
Muralidharan G, Yacobi A, Blotner S, Bryzinski B, Carver A, Simon J, Faulkner R
(1992) Pharmacokinetics of bisoprolol (B) and hydrochlorothiazide (HCTZ) after
administration of a combination tablet of B/HCTZ (1O/6.25mg) in fasted and
nonfasted healthy volunteers. Pharm Res 9 [Suppl]:S321
Nakamura H, Iwai N (1992a) Pharmacokinetic study on oral antibiotics in pediatrics.
III. A pharmacokinetic study on cefprozil in pediatrics. Jpn J Antibiot 45:1489-
1504
Nakamura H, Iwai N (1992b) Pharmacokinetic studies on oral antibiotics in pediatrics.
V. A pharmacokinetic study on cefdinir in pediatrics. Jpn J Antibiot 45:12-27
Nakashima M, Uematsu T, Oguma T, Yoshida T, Kimura Y, Konishi M (1993) Phase
I study of S-1108, a new ester-type oral cephem antibiotic. Chemotherapy 41
[Suppl1):109-125
Narang PK, Lewis RC, Bianchine JR (1992) Rifabutin absorption in humans: relative
bioavailability and food effect. Clin Pharmacol Ther 52:335-341
Nazareno LA, Holazo AA, Limjuco R, Massarella JW, Koss-Twardy S, Min B (1992)
The effect of food on absorption of dideoxycytidine (ddC) in HIV-positive pa-
tients. Pharm Res 9:S321
Neuvonen PJ, KivistO KT (1989) The clinical significance of food-drug interactions: a
review. Med J Aust 150:36-40
Neuvonen PJ, KivistO KT (1992) Milk and yoghurt do not impair the absorption of
ofloxacin. Br J Clin Pharmacol 33:346-348
Neuvonen PJ, KivistO KT, Lehto P (1991) Interference of dairy products with the
absorption of ciprofloxacin. Clin Pharmacol Ther 50:498-502
Nichols AI, Everitt D, Portelli S, Moore N, Law D, Yeulet S, Howland K, Audet P,
Rizzo S, IIson B, Jorkasky D (1992) The effect of food on the single dose pharma-
cokinetics of the leukotriene receptor antagonist SK&F 106203. Pharm Res 9
[Suppl):S309
Nilsen OG, Dale 0 (1992) Single dose pharmacokinetics of trazodone in healthy
subjects. Pharmacol Toxicol 71:150-153
Oguey D, Kolliker F, Gerber NJ, Reichen J (1992) Effect of food on the bioavailability
of low-dose methotrexate in patients with rheumatoid arthritis. Arthritis Rheum
35:611-614
Oguma T, Yamada H, Sawaki M, Narita N (1991) Pharmacokinetic analysis of
the effects of different foods on absorption of cefaclor. Antimicrob Agents
Chemother 35:1729-1735
Ohdo S, Nakano S, Ogawa N (1992) Circadian changes of valproate kinetics depending
on meal condition in humans. J Clin Pharmacol 32:822-826
Ohtsubo K, Fujii N, Higuch S, Aoyama T, Goto I, Tatsuhara T (1992) Influence of food
on serum ambenonium concentration in patients with myasthenia gravis. Eur J
Clin PharmacoI42:371-374
Olanoff LS, Walle T, Cowart TD, Walle KU, Oexmann MJ, Conradi EC (1986) Food
effects on propranolol systemic and oral clearance: support for a blood flow
hypothesis. Clin Pharmacol Ther 40:408-414
Oosterhuis B, Jonkman JHG (1993) Pharmacokinetic studies in healthy volunteers in
the context of in vitro/in vivo correlations. Eur J Drug Metab Pharmacokinet
18:19-30
Drug-Food Interactions Affecting Drug Absorption 119

Pan HY, Devault AR, Brescia D, Willard DA, McGovern ME, Whigan DB, Ivashkiv
E (1993) Effect of food on pravastatin pharmacokinetics and pharmacodynamics.
Int J Clin PharmacoI31:291-294
Palazzini E, Cristoformi M, Babbini M (1992) Bioavailability of a new controlled-
release oral naproxen formulation with and without food. Int J Clin Pharmacal
Res 12:179-184
Peck RW, Weatherley BC, Posner J (1993) The pharmacokientics of 882C, a thymidine
analogue with potent anti-VZV activity in healthy volunteers. Antiviral Res 20
[Suppl 1]:132
Pfeifer S (1993) EinfluB von Nahrung auf die Pharmakokinetik von Arzneistoffen.
Pharmazie 48:3-16
Pfeiffer A, Vidon N, Bovet M, Rongier M, Bernier 11 (1990) Intestinal absorption of
amiodarone in man. J Clin Pharmacol 30:615-620
Phelan MJI, Orme M, Williams E, Thompson RN (1991) Food has no effect on the
pharmacokinetics of methotrexate in patients with rheumatoid arthritis. Arthritis
Rheum 34 [Suppl]:S91
Pieniaszek HJ, Rakestraw DC, Schary WL, William RL (1991) Influence of food on the
oral absorption and bioavailability of moricizine. J Clin PharmacoI31:792-795
Plezia PM, Thornley SM, Kramer TH, Armstrong EP (1990) The influence of enteral
feedings on sustained-release theophylline absorption. Pharmacotherapy 10:356-
361
Plusquellec Y, Compistron G, Stevens S, Barre J, Jung J, Tillement JP, Houin G (1987)
A double-peak phenomenon in the pharmacokinetics of veralipride after oral
administration: a double site model for drug absorption. J Pharmacokinet
Biopharm 15:225-239
Prescott LF, Yoovathaworn K, Makarananda K, Saivises R, Sriwatanakul K (1993)
Impaired absorption of paracetamol in vegetarians. Br J Clin Pharmacol 36:237-
240
Ptachcinski RJ, Venkataramanan R, Rosenthal JT, Burckart GJ, Taylor RJ, Hakala
TR (1985) The effect of food on cyclosporine absorption. Transplantation 40:174-
176
Rapin J-R (1992) Effect of food on calcium antagonists pharmacokinetics. Pharm
Weekbl14 [Suppl E]:E5
Riva E, Gerna M, Latini R, Giani P, Volpi A, Moggioni A (1982) Pharmacokinetics of
amiodarone in man. J Cardiovasc Pharmacol 4:264-269
Roberts RJ, Leff RD (1989) Influence of absorbable and non absorbable lipids and
lipidlike substances on drug bioavailability. Clin Pharmacol 45:299-304
Robertson DRC, Higginson I, MacKlin BS, Renwick AG, Waller DG, George CF
(1991) The influence of protein containing meals on the pharmacokinetics of
levadopa in healthy volunteers. Br J Clin PharmacoI31:413-417
Roe AR (1989) Diet and drug interactions. Van Nostrand Reinhold, New York
Rolan PE, Mercer AJ, Westherley BC, Holdich T, Ridout G, Meire H, Posner J (1992)
Investigation of the factors responsible for a food-induced increase in absorption
of a novel antiprotozoal drug 566C80. Br J Clin Pharmacol 33:226P-227P
Rolan PE, Mercer AJ, Weatherley BC, Holdich T, Meire H, Peck RW, Ridout G,
Posner J (1994) Examination of some factors responsible for food-induced in-
crease in absorption of atovaquone. Br J Clin Pharmacol 37:13-20
Roller S, Lode H, Stelzer I, Deppermann KM, Boeckh M, Koeppe P (1992) Pharmaco-
kinetics of loracarbef and interaction with acetylcysteine. Eur J Clin Microbiol
Infect Dis 11:851-855
Roncari G (1993) Human pharmacokinetics of aniracetam. Drug Invest 5 [Suppl1 ]:68-
72
Roos RAC, Tijssen MAJ, Van der Velde EA, Breimer DD (1993) The influence of a
standard meal on Sinemet CR absorption in patients with Parkinson's disease.
Clin Neurol Surg 95:215-219
Rosenborg J, Nyberg L, Delen A-M (1991) Dietary habits do not influence the dosage
of Bambec® tablets. Eur Respir J 4 [Suppl 14]:434S
120 P.G. WELLING

Royer-Morrot M-J, Zhiri A, Jacob F, Necciari J, Lascombes F, Royer RJ (1993)

Influence of food intake on the pharmacokinetics of a sustained release formula-
tion of sodium valproate. Biopharm Drug Dispos 14:511-518
Ruhnke M, Bauer FE, Seifert M, Trautmann M, Hille H, Koeppe P (1993) Effects of
standard breakfast on pharmacokinetics of oral zidovudine in patients with AIDS.
Antimcrob Agents Chemother 34:2153-2158
Russell GAB, Martin RP (1989) Flecainide toxicity. Arch Dis Child 64:860-862
Sahai J, Gallicano K, Garber G, McGilvery I, Hawley-Foss N, Turgeon N, Cameron
DW (1992) The effect of a protein meal on zidovudine pharmacokinetics in HIV-
infected patients. Br J Clin Pharmacol 33:657--660
Saito A (1993) Effects of food, ceruletide and ranitidine on the pharmacokinetics of the
oral cephalosporin S-1108 in humans. Chemotherapy 39:374-385
Sakashita M, Yokogawa M, Yamaguchi T, Sekine Y (1991) Pharmacokinetics of
sparfloxacin in man. Yakubutsu Dotai 6:43-51
Sakashita M, Yamaguchi T, Miyazaki H, Sekine Y, Nomiyama T, Tanaka S, Miwa
T, Harasawa S (1993) Pharmacokinetics of the gastrokinetic agent mosapride
citrate after single and multiple oral administration in healthy subjects.
Arzneimelforschung 43:867-872
Sanchez N, Sheiner LB, Halkin H, Melmon KL (1973) Pharmacokinetics of digoxin:
interpreting bioavailability. Br Med J 20:132-134
Sandberg A, Ragnarsson G, Johnson UE, Sjogren J (1988) Design of a new multiple-
unit controlled-release formulation of metoprolol - Metoprolol CR. Eur J Clin
Pharmacol 33 [Suppl]:S3-S7
Sandberg A, Abrahamsson B, Regardh C-G, Wieselgren I, Bergstrand R (1990) Phar-
macokinetic and biopharmaceutic aspects of once daily treatment with metoprolol
CR/ZOK: a review article. J Clin PharmacoI30:S2-S16
Schaefer HG, Beermann D, Horstmann R, Wargenau M, Heibel BA, Kuhlmann J
(1993) Effect of food on the pharmacokinetics of the active metabolite of the
prodrug repirinast. J Pharm Sci 82:107-109
Segre G, Cerretani D, Moltoni L, Urso R (1992) Pharmacokinetics of rufloxacin in
healthy volunteers. Eur J Clin Pharmacol 42:101-105
Sekikawa H, Nakano M, Takada M, Arita T (1980) Influence of dietary components on
the bioavailability of phenytoin. Chern Pharm Bull (Tokyo) 28:2443-2449
Semple HA, Koo W, Tam YK, Ngo L-Y, Coutts RT (1991) Interactions between
hydralazine and oral nutrients in humans. Ther Drug Monit 13:304-308
Setnikar I, Arigoni R (1988) Chemical stability and mode of gastrointestinal absorption
of sodium monofluorophosphate. Arzneimittelforschung 38:45-49
Shah J, Fratis A, Ellis D, Murakami S, Teitelbaum P (1990) Effect of food and antacid
on absorption of orally administered ticlopidine hydrochloride. J Clin Pharmacol
30:733-736
Shelton MJ, Morse GD, Portmore A, Blum R, Saddler B, Reichman RC (1993)
Prolonged, but not diminished zidovudine (ZDV) absorption by food. Clin
Pharmacol Ther 53:207
Shinkuma D, Hamaguchi T, Kobayashi M, Yamanaka Y, Mizuno N (1990) Effects of
food intake and meal size on the bioavailability of sulpiride in two dosage forms.
Int J Clin Pharmacol 28:440-442
Shukla VA, Pittman KA, Barbhaiya RH (1992) Pharmacokinetics interactions of
cefprozil with food, propantheline, metoclopramide, and probenecid in healthy
volunteers. J Clin PharmacoI32:725-731
Shyu WC, Knupp CA, Pittman KA, Dunkle L, Barbhaiya RH (1991) Food-induced
reduction in bioavailability of didanosine. Clin Pharmacol Ther 50:503-507
Simon JA, Robinson DE, Andrews MC, Hildebrand JR III, Rocci ML, Blake RE,
Hogden GD (1993) The absorption of micronized progesterone: the effect of food,
dose proportionality, and comparison with intramuscular progesterone. Fertil
Steril 60:26-33
Smith HT, Jokubaitis LA, Troendle AJ, Hwang DS, Robinson WT (1993) Pharmaco-
kinetics of fluvastatin and specific drug interactions. Am J Hypertens 6:375S-382S
Drug-Food Interactions Affecting Drug Absorption 121

Sorgel F, Kinzig M (1993) Pharmacokinetics of gyrase inhibitors. I. Basic chemistry and

gastrointestinal disposition. Am J Med 94:44S-55S
Sun JX, Cipriano A, Chan K, Klibaner M, John VA (1994) Effect of food on the
relative bioavailability of a hypolipidemic agent (CGP 43371) in healthy subjects.
J Pharm Sci 83:264-266
Svensson CK, Edwards DJ, Mauriello PM, Barde SH, Foster AC, Lanc RA, Middleton
E Jr, Lalka D (1983) Effect of food on hepatic blood flow: implication in the food
effect phenomenon. Clin Pharmacol Ther 34:316-323
Tam YK, Kneer J, Dubach UC, Stoeckel K (1990) Effects of timing of food and fluid
volume on cefetamet pivoxil absorption in healthy normal volunteers. Antimicrob
Agents Chemother 34:1556-1559
Tanimura M, Kataoka S, Fujita Y (1991) Basic and clinical studies of sparfloxacin in
urinary tract infections. Chemotherapy 39 [Suppl 4]:523-530
Tay LK, Sciacca MA, Sostrin MB, Farmen RH, Pittman KA (1993) Effect of food on
the bioavailability of gepirone in humans. J Clin PharmacoI33:631-635
Taylor DM, Massey CA, Willson WG, Dhillon SA (1993) Lowered phenytoin serum
concentrations during therapy with liquid food concentrates. Ann Pharmacother
27:369
Terhaag B, Hrdlicka P, Gramatte T, Richter K, Feller K (1990) Zum EinfluB der
Nahrung auf die Pharmakokinetik von Diclofenac aus Rewodina-25-dragees. Z
Klin Med 45:443-446
Terhaag B, Gramatte T, Hrdlicka P, Richter K, Feller K (1991) The influence of food
on the absorption of diclofenac as a pure substance. Int J Clin PharmacoI29:418-
421
Thakker KM, Mangat S, Wagner W, Castellana J, Kochak GM (1992) Effect of
food and relative bioavailability following single doses of diclofenac 150mg
hydrogel bead (HGB) capsules in healthy humans. Biopharm Drug Dispos 13:327-
335
Thebault JJ, Montay G, Ebmeier M, Douin MJ, Millerioux L, Chassard D, Mignot A
(1990) Effect offood on the bioavailability ofthe new quinolone sparfloxacin. Proc
30th Interscience Conference on Antimicrob Agents Chemotherapy, Atlanta, Oct.
21-24, pp 294
Theodor RA, Weimann H-J, Weber W, Muller M, Michaelis K (1992) Influence of food
on the oral bioavailability of moxonidine. Eur J Drug Metab Pharmacokinet
17:61-66
Trovato A, Nuhlicek DN, Midtling JE (1991) Drug-nutrient interactions. Am Fam
Physician 44:1651-1658
Tsutsumi K, Nakashima H, Kotegawa T, Nakano S (1992) Influence of food on the
absorption of beta-methyldigoxin. J Clin PharmacoI32:157-162
Ueda Y, Matsumoto F, Imai T, Sakurai I, Takahashi T, Morita M (1993) Effect of
probenecid and a meal load on the pharmacokinetics of zidovudine. Chemo-
therapy (Tokyo) 41:499-503
Unadkat JD, Collier AC, Crosby SS, Cummings D, Opheim KE, Corey L (1990)
Pharmacokinetics of oral zidovudine (azidothymidine) in patients with AIDS
when administered with and without a high-fat meal. AIDS 4:299-232
Van Harten J, Van Bemmel P, Dobrinska MR, Ferguson RK, Raghoebar M (1991)
Bioavailability of fluvoxamine given with and without food. Biopharm Drug
Dispos 12:571-576
Van Peer A, Woestenborghs R, Heykants J, Gasparini R, Gauwenbergh G (1989) The
effects of food and dose on the oral systemic availability of itraconazole in healthy
subjects. Eur J Clin Pharmacol 36:423-426
Vance-Bryan K, Guay DRP, Rotschafer JC (1990) Clinical pharmacokinetics of
ciprofloxacin. Clin Pharmacokinet 19:434-461
Walter-Sack IE (1992) What is "fasting" drug administration? Eur J Clin Pharmcol
42:11-13
Walter-Sack IE, De Vries JX, Nickel B, Stenzhorn G, Weber E (1989) The influence of
different formula diets and different pharmaceutical formulations on the systemic
122 P.G. WELLING

availability of paracetamol, gallbladder size, and plasma glucose. Int J Clin

Pharmacol 27:544-550
Warneke G, Setnikar I (1993) Effects of a meal on the pharmacokinetics of fluoride
from oral monofluorophosphate. Arzneimittelforschung 43:590-595
Weber C, Roos B, Birnboeck H, Van Brummelen P (1993) Effect of food on the oral
bioavailability of the renin inhibitor remikirin (RO 42-5892). Br J Clin Pharmacol
36:177P-178P
Weimann H-J, Rudolph M (1992) Clinical pharmacokinetics of moxonidine. J
Cardiovasc Pharmac 20 [Suppl] 4:S37-S41
Welling PG (1977) Influence of food and diet on gastrointestinal drug absorption: a
review. J Pharmacokinet Biopharm 5:291-334
Welling PG (1989) Effects of food on drug absorption. Pharmacol Ther 43:425-441
Welling PG (1993) Necessity of food studies: implications of food effects. In: Midha
KK, Blume HH (eds) Bio-international: bioavailability, bioequivalence and phar-
macokinetics. Medpharm, Stuttgart, pp 211-221
Welling PG, Tse FLS (1982) The influence of food on the absorption of antimicrobial
agents. J Antimicrob Chemother 9:7-27
Welling PG, Elliott RL, Pitterle ME, Corrick-West HP, Lyons LL (1979) Plasma levels
following single and repeated doses of erythromycin estolate and erythromycin
stearate. J Pharm Sci 68:150-155
Wells TG, Sinaiko AR (1991) Antihypertensive effect and pharmacokinetics of
nitrendipine in children. J Pediatr 118:638-843
Welty DF, Siedlick PH, Posvar EL, Selen A, Sedman AJ (1994) The temporal effect of
food on tacrine bioavailability. J Clin Pharmacol 34:985-988
Wessels JC, Koeleman HA, Boneschans B, Steyn HS (1992) The influence of different
types of breakfast on the absorption of paracetamol among members of an ethnic
group. Int J Clin Pharmacol 30:208-213
Wilding IR, Hardy JC, Maccari M, Ravelli V, Davis SS (1991a) Scintigraphic and
pharmacokinetic assessment of a multiparticulate sustained release formulation of
diltiazem. Int J Pharm 76:133-143
Wilding IR, Hardy JG, Davis SS, Melia CD, Evans DF, Short AH, Sparrow RA, Yeh
KC (1991b) Characterization of the in vivo behavior of a controlled-release formu-
lation of levodopa (Sinemet CR). Clin NeuropharmacoI14:305-321
Wilding IR, Davis SS, Sparrow RA, Bloor JR, Hayes G, Ward GT (1992) The effect of
food on the in vivo behavior of a novel sustained release formulation of tiaprofenic
acid. Int J Pharm 83:155-161
Williams DB, O'Reilly WJ, Boehm G, Story MJ (1990) Absorption of doxycycline from
a controlled release pellet formulation: the influence of food on bioavailability.
Biopharm Drug Dispos 11:93-105
Williams L, Davis JA, Lowenthal DT (1993) The influence of food on the absorption
and metabolism of drugs. Clin Nutr 77:815-829
Williams PE, Harding SM (1984) The absolute bioavailability of oral cefuroxime axetil
in male and female volunteers after fasting and after food. J Antimicrob
Chemother 13:191-196
Willis JV, Kendall MJ, Jack DB (1981) The influence of food on the absorption of
diclofenac after single and mUltiple oral doses. Eur J Clin Pharmacol 19:33-37
Wilson CG, Washington N, Greaves JL, Washington C, Wilding IR, Hoadley T, Simms
EE (1991) Predictive modelling of the behavior of a controlled release buflomedil
HCl formulation using scintographic and pharmacokinetic data. Int J Pharm
72:79-86
Wix AR, Doering PL, Hatton RC (1992) Drug-food interaction counseling programs at
teaching hospitals. Am J Hosp Pharm 49:855-860
Wright CE (1992) The influence of meal timing on the kinetics (PK) and dynamics
(PD) of alprazolam sustained release tablets (ASR). Pharm Res 9 [Suppl]:S285
Yamaguchi T, Ikeda C, Sekine Y (1986) Intestinal absorption of a j3-adrenergic block-
ing agent Nadolol. II. Mechanism of the inhibitory effect on the intestinal absorp-
tion of Nadolol by sodium cholate in rats. Chern Pharm Bull (Tokyo) 34:3836-3843
Drug-Food Interactions Affecting Drug Absorption 123

Yogendran L, McLaughlin K, Vadas EB, Margolskee DJ, Bechard S, Hseih JY-K, Lin
C, Rogers JD, Sciberras DG, James 1M (1992) Comparative bioavailability of
different formulations of MK-679 an orally active LTD4 receptor antagonist. Br J
Clin PharmacoI34:169P-170P
Yong C-L, Yu D, Eden L, Eichmeyer L, Giessing D (1991) Effect of food on the
pharmacokinetics of oxybutinin in normal subjects. Pharm Res 8 [Suppl]:S320
Yu DK, Morrill B, Bhargava VO, Giesing DH, Weir SJ (1992) Effect of food
coadministration on Cardizem CDTM capsule bioavailability. Pharm Res 9 [Suppl]:
S323
Yuen KH, Deshmukh AA, Newton JM, Short M, Melchor R (1993) Gastrointestinal
transit and absorption of theophylline from a multiparticulate controlled release
formulation. Int J Pharm 97:61-77
Zimmermann T, Yeates RA, Laufen H, Pfaff G, Wildfeuer A (1994) Influence of
concomitant food intake on the oral absorption of two triazole antifungal agents,
itraconazole and fluconazole. Eur J Clin PharmacoI46:147-150
CHAPTER 4
Drug Interactions at Plasma
and Tissue Binding Sites
J.e. McELNAY

A. Introduction
The binding of drugs to plasma and tissue proteins is a major determinant of
drug distribution throughout the body. This binding has also a very important
effect on drug dynamics since it is only the free (unbound) drug which can
diffuse to, and interact with, receptor sites, i.e., bound drug is pharmacologi-
cally inactive. The relative ease of study in vitro of the binding of drugs in
human plasma and in separate protein fractions (mainly albumin) has led to an
extensive literature on the subject, with the influence of other drugs and/or
endogenous materials on drug binding being a common topic for investigation.
The relative difficulty in obtaining samples of viable human tissues has meant
that work on the binding of drugs to human tissues has been very limited by
comparison, although this may change with the more widespread use of tissue
culture techniques. It is as a result of this large body of largely in vitro data that
the topic of protein binding displacement interactions has gained prominence
as a possible mechanism of drug-drug interactions; however, as mentioned in
Chap. 1 of this volume, the importance of plasma protein binding displace-
ment interactions has been overestimated and overstated and only in very
specific cases do they result in adverse outcomes. In this chapter the influence
of plasma and tissue binding on drug kinetics is discussed to facilitate consid-
eration of the consequences of drug displacement from binding sites and their
possible clinical significance. Those cases in which displacement can lead to
clinically significant outcomes, which are much more restrictive than much of
the literature and popular belief amongst practitioners would suggest, are
considered in some detail. Before moving on the these issues, the proteins
involved in drug binding are described.

B. Proteins Involved in Drug Binding

The main binding proteins in plasma are albumin and lXt -acid glycoprotein.
Albumin consists of a single polypeptide chain and is present in plasma at a
concentration of 3S-4Sg/1 (approximates to O.6mM) in normal healthy indi-
viduals. Despite its large molecular weight, albumin is not retained exclusively
in the plasma compartment but is distributed extravascularly. Most of the
albumin that moves out of the vascular system is returned to the bloodstream
126 J.e. MCELNAY

via the thoracic duct (ROTHSCHILD et al. 1973). Approximately 13 g albumin is

synthesised (liver) and catabolised daily. Lowered serum albumin concentra-
tions can be precipitated by many factors, e.g., neoplastic disease, burns, and
lowered albumin concentrations are often seen in the elderly, particularly in
those elderly patients who are suffering from chronic disease. The binding of
albumin can be compromised by the presence of increased concentrations
of endogenous displacing agents, e.g., in kidney disease. Albumin, unlike lXt-
acid glycoprotein, which only binds basic drugs, can bind both acidic and
basic drugs. The binding of drugs by these plasma proteins is of course
dependent on the concentration of the protein in the plasma. Diseases/events
that can increase the concentration of lXt-acid glycoprotein will therefore result
in increased drug binding, e.g., myocardial infarction and Crohn's disease
(PIAFSKY et al. 1978).
As well as protein concentration, the other main determinants of the
amount of protein binding in plasma are the number of binding sites per
molecule of protein and the strength of binding between the drug and its
binding site. Two major binding areas have been identified on albumin, e.g.,
piroxicam has been shown to bind to the apazone locus (site I area) and to a
lesser extent to the diazepam site (site II) (BREE et al. 1990). The strength of
binding is measured in terms of the association constant between drug and
protein; the larger the association constant the more avid the binding (D'ARCY
and McELNAY 1982). A drug may have different classes of binding sites with
different affinities, e.g., tolbutamide has been estimated to have 2.27 primary
binding sites (with an association constant of 21.86 x 104 M-l) and 8.21 second-
ary binding sites (with an association constant of 1.71 x 102 M-l) per molecule
of albumin (CROOKS and BROWN 1974). If a drug has only one class of binding
site, a straight line plot is achieved in the graphical representation devised by
Scatchard (Fig. 1), which, having binding data over a range of protein and/or
drug concentrations, can be used to determine n (number of binding sites per
molecule of protein) and Ka (association constant). The graph becomes curved
if more than one class of binding site is involved and interpretation becomes
more difficult. The difficulties involved in interpreting binding data have been
addressed by PLUMBRIDGE et al. (1978).
There is much current interest in chirality of drugs in relation to interac-
tion with receptor sites and the production of different pharmacological
effects by different optical isomers of the same compound. Binding of drugs to
plasma proteins can also take place in a stereoselective manner. A recent
paper, for example, examined the protein binding of propranolol and
verapamil enantiomers in maternal and foetal serum. It was found that for
propranolol in maternal but not in foetal serum, the difference in binding of
the R- and S-isomers was significant, with the R1S ratio being significantly
larger in the mother than in the foetus. For verapamil, the difference in
binding of the R - and S-enantiomers was significant in both the mother and the
infant, but the R1S ratio was similar in mother and foetus (BELPAIRE et al.
1995).
Drug Interactions at Plasma and Tissue Binding Sites 127

... .
.

...

Fig. 1. Plot devised by SCATCHARD (1949) for drug/macromolecule interaction for a

single class of binding site (r moles of drug bound per mole protein; D f molar free drug
concentration; n number of primary binding sites per molecule protein; Ka association
constant of drug for binding sites). If more than a single class of binding site is involved
in the binding interaction, the plot will be curved

The protein binding of drugs in tissues also primarily involves albumin

(JUSKO and GRETCH 1976). Other agents are, however, involved, e.g., digoxin
binds strongly to cardiac muscle (DOHERTY et al. 1967), bilirubin binds
with ligandin (DAVIS and YEARY 1975), certain cytotoxics bind with DNA
(ALLEN and CREAVEN 1974) while many drugs are bound by melanin
(SALAZAR-BoOKAMAN 1994). In addition, some drugs are bound to red
blood cells. Recent in vitro binding studies, for example, have indicated
the presence of at least three kinds of binding sites on red blood cells
for dorzolamide, a novel topical carbonic anhydrase inhibitor (HASEGAWA et
al. 1994).

c. Influence of Plasma and Tissue Binding

on Drug Kinetics
Both plasma and tissue binding of drugs, via their influence of maintaining a
high concentration gradient of free drug between the lumen of the intestine
and plasma post administration of a drug orally, will favour a more rapid
absorption of drugs from the intestine (or indeed injection site). These effects
are, however, very marginal in comparison to the influence of protein binding
on drug distribution and elimination.
Regarding drug distribution, plasma and tissue protein binding have
opposing effects. Since plasma proteins are largely retained within the plasma
128 J.C. MCELNAY

compartment, drugs which are highly bound to plasma proteins tend to have
low apparent volumes of distribution (Vd). Warfarin, the best-known example
of a drug which is highly bound to plasma albumin (>99%), has a low Vd (0.111
kg), indicating poor diffusion into the tissues (TILLEMENT et al. 1978). Drugs
which are highly bound to tissues tend to have a high Vd. This, of course, will
depend on the type of tissue to which the drug is bound and its abundance in
the body. For drugs which are highly bound to both plasma and tissues, the Vd
will depend on the relative binding in both sites. With tricyclic antidepressants
and phenothaizines, for example, almost all of the drug in plasma is bound to
albumin; however, due to extensive tissue binding of these agents, the circulat-
ing drug in plasma represents only a small fraction of the total drug in the body
(KOCH-WESER and SELLERS 1976). The equilibrium which exists between
tissue and plasma binding is represented in Fig. 2. As mentioned earlier, only
unbound drug is available to diffuse to receptor sites and exert therapeutic
effects.
The next phase in the pharmacokinetic process is drug elimination, involv-
ing primarily the liver and kidney. It is obvious that if a drug is highly bound
to tissues, it will be protected against elimination since much of the drug
will be outside the plasma compartment and will therefore not be "delivered"
to the kidney or liver for elimination. In addition, the extent of plasma
protein binding can have marked effects on drug clearance by these organs.
Consider first drug metabolism. For most drugs, plasma protein binding is
protective in that the affinity for plasma binding sites exceeds that for
metabolising systems. It is therefore only free (unbound) drug that is
metabolised during passage through the liver (or other metabolic site). Such
drugs would be considered to have a low hepatic extraction ratio (low
tendency to be removed from the blood during its passage through the liver).
On the other hand, if a drug has a high extraction ratio, the attraction to
metabolic enzyme systems in the liver will overcome the binding affinity for
plasma proteins and bound drug will be stripped from the binding sites to
undergo metabolic change (McELNAY and D'ARCY 1983). This will often
involve active uptake of the drug into liver cells. In the case of high extraction
drugs, plasma protein binding will facilitate drug metabolism by maintaining
high total drug concentrations in the plasma (via decreased drug distribution)
and will essentially act as a carrier system, taking the drug to its site of
destruction. The metabolism of highly extracted drugs, e.g., propranolol, are
dependent upon liver blood-flow (ROWLAND 1980).
A similar situation pertains to the kidney. Glomerular filtration, being
a passive process, allows only free (unbound) drug to pass through the
glomerulus in the normal healthy kidney, since protein molecules are too
large to be filtered. Plasma protein binding therefore protects a drug against
elimination in the kidney by glomerular filtration. As the filtrate passes down
the renal tubules, back diffusion of drug occurs when the concentration of
free drug in the renal tubule (due to water reabsorption) exceeds the free
Drug Interactions at Plasma and Tissue Binding Sites 129

Plasma Tissues

Bound Bound
drug drug

II
Free .c::.
1l
Free
drug drug
/'

Fig.2. Distribution of drug between plasma and tissue compartments. Only free (phar-
macologically active) drug can diffuse out of the plasma; the equilibrium would there-
fore move towards the right after displacement from plasma binding and towards the
left after displacement from tissue binding sites. Free drug concentrations in the plasma
and tissue compartments are equal. Quantities of drug bound will depend on the
amount of binding materials present in both plasma and tissues and their respective
binding capacities. (After TILLEMENT 1980)

drug concentration in plasma. High binding in plasma helps maintain a

high concentration gradient for the back diffusion of drugs which are lipid
soluble at the pH of tubular filtrate, again helping to pTevent drug elimination.
Although glomerular filtration is the main mechanism of drug elimination
via the kidneys, a number of drugs, e.g., penicillins, are actively secreted
into tubular urine. As in the case of high extraction in the liver, the affinity
for these active processes overcomes the binding affinity in plasma. Therefore
plasma protein binding tends to facilitate elimination by limiting drug
distribution, thereby allowing more drug to be presented to the kidney for
elimination.

D. Displacement of Drugs from Binding Sites

In the case of plasma protein binding, most displacement of drugs from plasma
binding sites is via a direct competition of the two drugs involved for the same
binding sites. Drugs with a higher association constant will displace drugs
with a lower association constant at common binding sites. Probably the
best-known example of such a displacement interaction is between warfarin
and phenylbutazone, with phenylbutazone successfully displacing warfarin
130 J.e. MCELNAY

from its binding. In this type of interaction the phenylbutazone is termed the
"displacer" drug and warfarin the "displaced" drug, the latter term being
usually reserved for the drug with the higher potential to give rise to adverse
effects (KOCH-WESER and SELLERS 1976). Since the displacement is competi-
tive, the binding of both drugs will be decreased. For substantial displacement
to take place, the displacer must occupy the majority of the available binding
sites, thereby substantially lowering the binding sites available to the primary
drug (ROWLAND and TOZER 1995). For this to take place the molar concentra-
tion of drug in plasma must exceed the molar concentration of albumin
(approximately 150.ug/ml for a drug with a molecular weight of 250). As a
result of these prerequisites relatively few drugs are commonly referred to as
displacers, e.g., salicylic acid, valproic acid and phenylbutazone. Since the
molar concentration of lXt-acid glycoprotein is 9-23.uM in plasma, lower molar
concentrations of basic drugs will be required to fulfil the latter displacer
criterion (ROWLAND and TOZER 1995). In certain cases the parent drug may not
itself be involved in a displacement interaction, but rather a metabolite may
act as the displacer, e.g., trichloracetic acid displaces warfarin while the parent
drug chloral hydrate does not (SELLERS and KOCH-WESER 1970).
As well as the competitive displacement interactions, the binding of one
drug to, e.g., albumin, can give rise to changes in the conformation of the
protein and thereby change the shape of the binding sites for a second drug.
This normally gives rise to a decreased binding of the second drug although
occasionally with allosteric effects the binding affinity of the second drug is
increased. Aspirin influences the binding of certain drugs by this non-competi-
tive mechanism, e.g., flufenamic acid and phenylbutazone, via permanently
acetylating lysine residues of albumin (PINKARD et al. 1973). The alkylating
agents carmustine and mechlorethamine give rise to decreased binding of
penbutolol to lXt -acid glycoprotein. Dialysis against saline for 24 h does not
restore the free fraction of penbutolol. The authors who conducted this latter
research concluded that treatment of cancer patients with alkylating agents
could alter the serum proteins and modify their binding capacity and that this
should be taken into account when co-administering other basic drugs, e.g.,
methadone (AGUIRRE et al. 1994).
The study of the binding of drugs to tissue components had been much
more limitied that those involving plasma; however, it is likely that both
competitive and non-competitive binding displacement also takes place at
tissue binding sites.

E. Therapeutic Consequences
of Plasma Binding Displacement Drug-Drug Interactions
The interest in plasma protein binding displacement interactions was first
aroused by the well-known interaction between warfarin and phenylbutazone
(AGGELER et al. 1967), which results in a marked, and clinically significant,
Drug Interactions at Plasma and Tissue Binding Sites 131

increased prothrombin time. It was correctly demonstrated that phenyl-

butazone displaced warfarin from its plasma binding sites (albumin); however,
it was incorrectly assumed that the plasma binding displacement was the
underlying mechanism of the interaction. Similarly, the increased pharmaco-
logical effect of tolbutamide in the presence of sulphonamides (CHRISTENSEN
et a1. 1963) was attributed to plasma binding displacement of tolbutamide by
the sulphonamide. Again, if one examines the binding of these latter two drugs
in vitro, there is undoubted displacement of the tolbutamide. If, for example,
one places both sets of drugs together with plasma in an equilibrium dialysis
cell, the free concentration and free fraction (free drug/total drug) of warfarin
and tobutamide will be elevated, with corresponding decreases in the percent-
age of drug bound to the albumin component in plasma. In these early days
plasma binding displacement was heralded as an important interaction mecha-
nism, and the relative ease with which binding characteristics of drugs both
singly and in combination could be determined led to many papers being
published on drug-drug displacement interactions.
The problem with these data, however, was that the work was carried out
in vitro and whereas displaced drug cannot distribute further than the confines
of a dialysis cell, in the in vivo situation, displaced drug can distribute out of
the plasma compartment into the relatively large "sink" of tissues. This moves
the equilibrium situation presented in Fig. 2 in an anticlockwise direction. In
addition, immediately after displacement has taken place, increased amounts
of free drug will be presented to sites of elimination (liver and kidney). These
compensatory effects of redistribution and increased elimination of free drug
(assuming the drug has a low extraction ratio) lead to only a transient increase
in free (pharmacologically active) drug in the plasma, whilst at equilibrium
(steady-state) the free concentration ofthe drug will equate to pre-interaction
levels. In addition, since distribution of drug from plasma to its receptor is
often not instantaneous, but via a first-order process with comparable rate to
general drug distribution, this will also reduce the extent of the transient
increase in effect of the displaced drug. If the displacer drug is given orally,
rapid compensatory effects may prevent even the transient increases in free
drug concentration from taking a place. The displacement does, however, lead
to important changes in the displaced drug's apparent volume of distribution
(this will be increased), the total concentration of drug in plasma (decreased;
Fig. 3) and the free fraction of drug in plasma (increased). This means that the
same pharmacological effect will be achieved from a reduced total (free and
bound) serum concentration of the drug. This type of effect has been clearly
demonstrated in research which examined the effects of sodium salicylate on
the elimination of indomethacin in dogs. Indomethacin was administered in-
travenously (bolus followed by an infusion) to yield steady-state plasma levels
in the dogs. This was followed by administration of sodium salicylate either
intravenously or via a duodenal fistua1. Sodium salicylate administration gave
rise to an immediate fall in indomethacin plasma concentrations by some
60%-70%. Total systemic clearance and biliary excretion of indomethacin
132 J.e. MCELNAY

c:
Displacing
.~
~
Agent
2
C
cu
u
c:
o
U
I
o
Iii L..
..2
a..
o
....o
e
0>
o
.....J

TIME
Fig. 3. Typical plasma concentration-time profile for a drug interaction involving
plasma protein binding placement. An almost immediate drop in total plasma concen-
tration (due to an increased Vd) is seen. Assuming clearance is unchanged, elimination
half-life in this case is shown to be increased (decreased slope of elimination curve).
(After D'ARCY and McELNAY 1982)

were increased as was the Vd, while plasma protein binding was decreased
(BECK et al. 1990).
If follows therefore that interactions which have in the past been attrib-
uted solely to plasma binding displacement must have a different underlying
mechanism. The best example of this is the now familiar warfarin-
phenylbutazone interaction. As long ago as 1974, LEWIS et al. discovered a
metabolic mechanism for this interaction. Phenylbutazone was shown to ste-
reo-selectively inhibit the metabolism of S-warfarin (warfarin is available
commercially as a racemic mixture and S-warfarin is five to six times more
potent than R-warfarin; O'REILLY 1971). There was a corresponding increase
in the elimination of R-warfarin. The increased proportion of the more potent
S-warfarin, however, gave rise to the increased anticoagulation observed.
Azapropazone is another example of a non-steroidal anti-inflammatory agent
which gives rise to markedly increased prothrombin times when administered
together with warfarin (POWELL-JACKSON 1977). Although azapropazone is a
potent displacer of warfarin from plasma binding sites (McELNAY and D'ARCY
1980), this displacement cannot explain the increased pharmacological
effect seen, and, although it has never been investigated, it would appear
that a metabolic effect similar to that with phenylbutazone takes place
(azapropazone has structural similarities to phenylbutazone). (Although many
other non-steroidal anti-inflammatory agents may not increase prothrombin
time, it should be borne in mind that they can cause gastrointestinal bleeding,
which can be exacerabated in anticoagulated patients.)
Drug Interactions at Plasma and Tissue Binding Sites 133

Other interactions which were initially described as being due to plasma

binding displacement, and which have now been demonstrated to have
other mechanisms responsible for the interaction, include: warfarin!
sulphamethoxazole and sulphinpyrazone - inhibition of metabolism
(O'REILLY 1980; TOON et al. 1986); methotrexate/salicylate - inhibition of
renal clearance (LIEGLER et al. 1970; STEWART et al. 1991); phenytoin/valproate
- inhibition of metabolism (PERUCCA et al. 1980) and tolbutamide/
sulphonamides - inhibition of metabolism (POND et al. 1977).
An interesting example of decreased plasma protein binding and
decreased clearance was described by FERRAZZINI et al. (1990). This involved
the interaction between trimethoprim-sulphamethoxazole and methotrexate
in children with leukaemia. The free fraction of methotrexate was increased
in all patients (mean 37.4%-52.2%) with concomitant administration of
trimethoprim-sulphamethoxazle. Although plasma clearance did not change
significantly, renal clearance of free methotrexate decreased significantly. The
changes in protein binding and renal clearance of methotrexate resulted in a
mean 66% increase in the systemic exposure to the drug. The authors
suggested that this may exaplain the myelotoxicity often observed when the
drugs are administered together. Work in the rabbit suggests that a similar
interaction involving both decreased protein binding and renal clearance
when ketoprofen is coadministered with methotrexate (PERRIN et al. 1990).
Although decreased elimination is undoubedly the underlying cause of
the interactions, plasma protein binding displacement tends to complicate
the interpretation of the findings with regard to measured serum drug
concentrations.
Although in most cases plasma binding displacement interactions per se
do not give rise to clinically significant interactions, there are three main
exceptions to this general rule:

I. Rapid Intravenons Infusion of Displacing Agent

If the displacing agent is given rapidly as a bolus injection, due to rapid
displacement of the displaced drug, the free concentration could increase
dramatically, and, since the compensatory mechanisms may take a short time
to reduce the free drug concentration to pre-interaction levels, the displaced
drug could reach receptors or enter compartments that it would not normally
reach. The best-known example of such a displacement interaction involves
the intravenous administration of sulphonamides to neonates. This results in
the displacement of plasma-bound bilirubin, which in turn enters the CSF,
resulting in kernicterus. Regarding drug-drug interactions, although conceiv-
able, such interactions are unlikely to arise because the displaced drug
would rapidly move down the newly created unbound concentration gradient
into the large tissue water space and because intravenous injections are not
normally administered rapidly in practice due to the adverse reactions often
134 J.e. MCELNAY

experienced (ROWLAND and TOZER 1995). A case report of an interaction in

which this mechanism may have been involved has, however, recently been
published. KISHORE et al. (1993) described a case of acute verapamil toxicity
in a patient with chronic verapamil toxicity, possibly precipitated by the
intravenous administration of the highly protein bound drugs ceftriaxone
and clindamycin. Administration of the ceftriaxone and clindamycin led to
complete heart block, requiring cardiopulmonary resuscitation and insertion
of a temporary pacemaker. The patient reverted to spontaneous sinus rhythm
some 16h later. The authors suggested that ceftriaxone, clindamycin or both
agents may have precipitated the toxicity via displacing verapamil from its
protein binding sites.

II. Parenteral Administration of Displaced Drug

Having High Extraction Ratio
Consider the case of a drug with a high extraction ratio which is administered
parenterally. Following an interaction that leads to displacement from plasma
protein binding sites, the drug will distribute into the tissues and other body
fluids, resulting in an initial decrease in plasma total drug concentrations.
However, clearance of the displaced drug will remain unchanged and unless
the dose of the drug is decreased the drug will begin to accumulate in the body
so that the total (free + bound) drug concentration in plasma will return to the
same value as prior to the interaction. (Steady-state plasma concentration
of total (free + bound drug) equals rate of drug administration divided by
clearance.) This will result in increased free drug concentrations in the plasma
and therefore an enhanced therapeutic effect. Examples of drug interactions
involving this latter mechanism are difficult to find - one drug which fulfils the
criteria of having a high extraction ratio and being administered repeatedly by
intravenous injection is lignocaine (ROLAN 1994). If the drug has a high extrac-
tion ratio in the liver, and is administered orally, due to decreased oral
bioavailability as a result of a higher first-pass effect, the drug would not be
expected to accumulate in the body and any effect of displacement on free
drug concentrations at steady state is likely to be small (ROWLAND and TOZER
1995). However, if the high extraction ratio pertains to drug elimination in the
kidney, a build-up of free drug concentration in the plasma after displacement
could be anticipated even after oral administration (McELNAY and D'ARCY
1983).

III. Therapeutic Drug Monitoring and Drug Displacement

from Plasma Binding Sites
For drugs with a narrow therapeutic range (narrow concentration range be-
tween therapeutic and toxic serum concentrations), to assist in patient man-
agement, therapeutic drug monitoring (TDM) is often performed. TDM
involves the measurement of drug concentration in plasma or whole blood
Drug Interactions at Plasma and Tissue Binding Sites 135

and adjustment of dosage based on measured drug concentrations. Total

concentration of drug is the normal measurement made; it is clear from Fig. 3
that this concentration will fall after drug displacement and redistribution has
taken place. The normal response to a decreased plasma drug concentration
during therapeutic drug monitoring would be to increase the dose of drug or
decrease the dosing interval. This would obviously be inappropriate since
the free drug concentration remains at pre-interaction levels. Increased
dosage would lead to drug accumulation and increased steady-state free
drug concentration with attendant toxicity. Clinicians therefore must take
displacement interactions into account when performing therapeutic drug
monitoring. A possible solution is to monitor free drug concentrations in
the plasma rather than total drug concentrations. Alternatively saliva drug
concentrations of the drug of interest can be measured since these correlate
with free drug concentrations (LINDUP 1975; McAULIFFE et al. 1977). An
example of this approach has also been suggested by TRNAVSKA et al. (1991),
who examined the kinetics of lamotrigine using both plasma and saliva
monitoring in patients undergoing long term antiepileptic therapy. The
authors suggested that salivary monitoring represented a non-invasive
alternative in therapeutic drug monitoring. These approaches are, however,
seldom used clinically.
An example of a drug interaction involving drugs which require therapeu-
tic drug monitoring is that between phenytoin and sodium valproate. Much
attention has been focused on this interaction. FRIEL et al. (1979), for example,
found that the mean total phenytoin plasma concentrations declined from 19.7
to 15.3 ,ug/ml when valproic acid was added to the regimen of 12 outpatients. In
20 patients receiving both drugs routinely, the median phenytoin free fraction
was 15.8%, which contrasted with a much lower free fraction value (9.1 %) in
40 patients who were receiving either phenytoin alone or phenytoin and
phenobarbitone. The authors suggested that the fall in total concentrations of
phenytoin could lead to inappropriate adjustment of phenytoin dosage. Theo-
retical consideration of the interaction by KOBER et al. (1978) led these authors
to conclude that valproic acid, with its high therapeutic plasma concentrations,
would increase the free fraction of phenytoin by 100%. PISANI and DIPERRI
(1981) found a significant fall in plasma and a concurrent significant increase in
salivary phenytoin when valproate (200mg intravenously) was administered
to ten epileptic patients treated long term with phenytoin. KNOTT et al. (1982)
in a much larger study involving measurement of salivary and plasma
phenytoin in 42 epileptic patients, indicated that saliva phenytoin concentra-
tions bore the same close relationship to unbound therapeutically active
phenytoin in patients receiving both drugs as it did in patients receiving
phenytoin alone, whereas total plasma phenytoin concentrations did not.
In a more recent study, MAy and RAMBECK (1990) examined the serum
concentrations of phenytoin in 28 epileptic patients who were treated with
phenytoin and valproate and in 15 patients who were treated with phenytoin
in the absence of valproate. The results indicated that there were greater than
136 J.e. MCELNAY

normal fluctuations in serum phenytoin concentrations when valproate was

coadministered. It was, however, noted that serum fluctuations in the concen-
trations of free (unbound) phenytoin did not differ between the two treatment
groups. Samples were taken at 0800, 1100, 1400, 1700, and in part at 2000
hours.

F. Therapeutic Consequences
of Tissue Binding Displacement Interactions
Displacement of drugs from tissue binding sites gives rise to the opposite effect
to that of plasma binding displacement, i.e. the equilibrium depicted in Fig. 2
moves in a clockwise direction, giving rise to redistribution of drug from the
tissues back into the plasma compartment. The plasma compartment, being
relatively small in comparison to the tissue compartment, may be unable to
buffer the impact of increasing amounts of free drug and therefore adverse
sequelae may be more likely. Since increased free drug is forced into the
plasma compartment, the total (free + bound) drug concentration will initially
be increased (Fig. 4) and there will be a reduction in the apparent volume of
distribution of the displaced drug. Elimination mechanisms will of course
come into play and the rate of elimination of the displaced drug in, for
example, the liver and the kidney will increase due to increased concentrations
of free drug. Assuming binding within the plasma remains unaffected and drug
clearance remains the same, both the mean total and the mean free drug
concentrations will re-equilibrate to their original pre-interaction values. This
re-equilibration, unlike the case with plasma binding displacement in which
redistribution into the "sink" of the tissues takes place very quickly, will
require some four to five half-lives. Therefore for some drugs with a long
half-life, mean plasma concentrations of free drug may remain elevated for
prolonged periods. Although clearance will not be effected, the reduction
in Vd will result in a decrease in the elimination half-life, which in turn will
lead to greater than normal fluctuations in plasma levels of the displaced
drug during a dosing interval. A major drug interaction can take place if
the displacing drug also inhibits the elimination of the displaced agent as
occurs in some of the known interactions which occur by this mechanism. The
best-known interactions that involve tissue displacement are with the cardiac
glycoside, digoxin. The most extensively studied interaction involves digoxin
and quinidine (Fig. 5). Administration of quinidine to a patient stabilised on
digoxin leads, as expected, to an increased digoxin concentration in plasma. In
one clinical trial, serum digoxin concentrations increased in 25 of 27 subjects
during co-administration of quinidine from a mean of 1.4ng/ml to 3.2ng/ml.
Signs of digoxin toxicity (anorexia, nausea and vomiting) were noted in 16
patients (59%) during concomitant quinidine therapy (REIFFEL et al. 1979).
The data presented in Fig. 5 indicate clearly that the volume of distribution
of digoxin is decreased during concomitant quinidine treatment - mean
Drug Interactions at Plasma and Tissue Binding Sites 137

c
.2
~c
II)
u
c
0

t
u
0
~
0
a: D isplocing
0 Agent
"0
~

0
OJ
0
-'

TIME

Fig.4. Typical plasma concentration time profile for a drug interaction involving tissue
binding displacement. An almost immediate increase in total plasma concentration
(due to a decreased Vd) is seen. Elimination half-life is shown to be increased. This is
because of increased amounts of both free and bound drug being presented at sites of
elimination. (After D' ARCY and McELNAY 1982)

40.0
30.0

20.0

10.0

5.0

10 20 30 40 50 60 70 80 90
Time (hours)

Fig. 5. Semilogarithmic plot of plasma concentration of digoxin as a function of time

after the intravenous administration of digoxin before quinidine treatment (o-D) and
during quinidine treatment (e-e) in one subject. (After HAGER et al. 1979)
138 J.e. MCELNAY

values for Vd were 1O.871Jkg for digoxin alone and 7.351/kg for digoxin
administration to six patients receiving quinidine (HAGER et al. 1979). These
authors suggested that the results could be explained by the displacement
of digoxin from tissue binding sites by quinidine. Renal clearance of digoxin
is also decreased by quinidine, probably due to inhibition of renal tubular
secretion, and therefore this compensatory mechanism is compromised, a
situation which will further exacerbate the outcome of the interaction. It is
obviously necessary to carefully monitor serum digoxin concentrations during
concomitant quinidine treatment and, indeed, DOERING (1979) has suggested
that digoxin dosage should be halved during quinidine therapy, as on average
digoxin concentration doubles during the interaction. There is, however, no
substitute to concentration monitoring since there is great variability in the
magnitude of the quinidine-induced effects, e.g., one patient may have no
induced increase in digoxin serum concentration while another may have a
fivefold increase (BIGGER 1979). Since discovery of the interaction involving
quinidine, several other structurally related drugs have been shown to be
involved in a similar interaction with digoxin. These drugs include the other
antiarrhythmics amiodarone and propafenane. Disopyramide is free from
interaction with digoxin and could be used to replace quinidine in patients
receiving digoxin, since it has similar properties and indications (KOCH-WESER
1979). Choroquine, an antimalarial drug, which is also used as a disease
modifying agent in the treatment of rheumatoid arthritis, has been shown to
increase digoxin plasma levels in dogs, presumably via a tissue binding
displacement interaction (McELNAY et al. 1985).
An interaction involving both displacement at plasma and tissue binding
sites will have an even greater potential for adverse effects since free drug
concentrations could be elevated to an even more marked extent.
Sulphaethidole, for example, may displace dicloxacillin at both sites (DE
SANTE et al. 1980).

G. Displacement of Drugs from Binding Sites

by Endogenous Materials
The best-known example of a drug-drug interaction, which also involves en-
dogenous materials, involves the parenteral anticoagulant heparin (DESMOND
et al. 1980). These authors have shown that heparin may cause a 150%-250%
increase in the measured free fraction of diazepam, chlordiazepoxide and
oxazepam, but no change in that of lorazepam, in normal non-fasting subjects.
The changes were noted within 90s of administration of 100 units of heparin;
binding returned to baseline in 30-45 min. The authors reported a twofold
increase in free fatty acid concentrations post administration of heparin and
that these endogenous materials were responsible for the decreased binding of
the affected benzodiazepines. Since sampling canulae are commonly flushed
Drug Interactions at Plasma and Tissue Binding Sites 139

with heparin to keep them patent during clinical trials, such an effect
could give rise to erroneous data. Heparin, for example, has also been shown
to alter the plasma binding of phenytoin (CRAIG et al. 1976), quinidine
(KESSLER et al. 1979), propranolol (WOOD et al. 1979a,b) and warfarin
(CHAKRABARTI 1978).
Although initially the displacement via endogenous fatty acids appeared
to be an in vivo effect, a careful study by GIACOMINI et al. (1980) indicated that
the displacement takes place after sample collection. Via in vitro hydrolysis of
plasma triglycerides by lipases, which are released and activated after in vivo
administration of heparin, free fatty acid concentrations in blood samples are
increased. The problem can largely be overcome by the use of a lipase inhibi-
tion in the sample collection vial and it is recommended that this precaution be
taken when performing protein binding analyses with drugs which can be
displaced by free fatty acids.

H. Disease States and Altered Plasma Protein Binding

Endogenous materials, which can act as displacing agents, can also be elevated
in a number of disease states. This phenomenon will give rise to essentially
the same pharmacokinetic picture as is seen in the presence of exogenously
administered displacing drugs. Free fatty acid concentrations are elevated,
for example, in stress or hyperlipidaemia, bilirubin in jaundice (BARRE et
al. 1983) and 2-hydroxybenzoylglyine in uraemia (FISET et al. 1986). In
insulin-dependent diabetes mellitus (IDDM) the increased serum concentra-
tions of free fatty acids gives rise to decreased binding of valproic acid (Fig. 6;
GRAINGER-RoUSSEAU et al. 1989). This is not surprising since valproic acid
itself is a fatty acid and is therefore involved in competition with endogenous
fatty acids for albumin binding sites.
Other factors which can give rise to changed binding in disease, and which
can obviously complicate drug-drug binding interactions, include alteration in
serum albumin concentrations (as in renal and hepatic disease), alteration
in albumin distribution due to an increased unidirectional permeability of
capillaries and impaired return of albumin to the plasma compartment due to
impaired drainage of lymph vessels (BARRE et al. 1983). Conformational
changes in binding protein molecules resulting in changes in drug binding
affinity can also take place, e.g., via glycosylation of albumin (SHAKLAI et
al. 1984). An increased concentration of a)-acid glycoprotein induced by
body stress, e.g., myocardial infarction or inflammation (PIAFSKY et al. 1978;
PIAFSKY 1980), can lead to increased binding of basic drugs. Increased plasma
concentrations of this latter protein are also seen in pregnancy. Marked
decreases in albumin concentrations can be found in a range of debilitating
disease states including burns, cancer, cardiac failure, cystic fibrosis, renal
failure, nutritional deficiency, liver diseases and post surgery (TILLEMENT et al.
140 J.C. McELNAY

40

E % Free 100M
35
... % FreeCtri

E
30

Q)
25
~
~
0
:Q
u 20
ro
u
'0
0..
<ii
>
15
,... ...
... ...
10
...
5

o ~~~ __ ~~~ __ ~~-L~~~~~ __~~

20 25 30 35 40 45 50 55
Serum albumin concentration (giL)

Fig. 6. Scattergraph of valproic acid % free vs serum albumin concentration in serum

from insulin-dependent diabetes mellitus patients and age-matched controls (total
serum concentration 90.ug/ml). (After GRAINGER-RoUSSEAU et al. 1989)

1978; 0IE 1986). All these conditions will complicate drug-drug interactions at
plasma binding sites. DASGUPTA et al. (1994) found that endogenous com-
pounds present in uraemic serum led to a decreased binding of valproate, but
that the displacement of valproate by salicylate was less marked in uraemic
serum when compared with normal serum. A review of the influence of disease
on plasma protein binding has been presented by GRAINGER-RoUSSEAU et al.
(1989) and a summary table from that review is presented as Table 1. In the
present context, this table serves two purposes; first it draws attention to
specific drug-disease interactions involving plasma protein binding and sec-
ondly data on the free fraction of a range of drugs which are bound to plasma
proteins are presented. The latter information is important bearing in mind
that clinically significant drug-drug displacement interactions in plasma usu-
ally involve drugs which are greater than 90% bound to plasma proteins at
normal therapeutic concentrations. Although pregnancy is not a disease, it has
been included in Table 1 since binding of some drugs will change during
pregnancy. It has recently been suggested, for example, that during pregnancy
and the 1st month after delivery, both total and free valproate serum concen-
trations should be closely monitored to determine the lowest effective dose
(JOHANNESSEN 1992).
Drug Interactions at Plasma and Tissue Binding Sites 141

Table 1. Influence of diseases on drug plasma protein binding

Drug Disease % Free or free Significant Clinical significance! References
fraction (control difference comment
vs. patient)

Carbamazepine Alcoholic liver 22.7% vs. 19.5% Y Not clinically significant. BARRUZI
disease, renal failure, Caution in interpreting et a1. 1986
rheumatoid arthritis, total drug concentrations.
ulcerative colitis Correlation with AGP
shown
Diazepam Early pregnancy 1.8% vs. 1.9% N Large V d and therapeutic PERUCCA
Mid pregnancy 1.8% vs. 2.1 % Y index means change is not et a!. 1981
Late pregnancy 1.8% vs. 2.6% Y clinically important in
chronic use. If used in
status epilepticus may have
potentiated action in late
pregnancy due to increased
free fraction
Uraemia 2.9% vs. 4.5% Not indicated. Site II not CALVO
affected by carbamylation. et al. 1982
Endogenous binding
inhibitors implicated
Uraemia 1.64 % vs. 3.23 % Y Not indicated, but an GROSSMAN
Kidney transplant 1.50% vs. 2.11 % Y important difference et a!. 1982
Nephrotic syndrome 1.60% vs. 3.55% Y implicated when
interpreting kinetic data
with regard to type of renal
renal disease and the
protein involved in binding
Chronic cardiac "=1 % VS. 1.5% N Does not alter diazepam FiTCHL
failure binding site affinity et al. 1983
Acute uraemia 2% VS. 6% Y Large therapeutic index TJULA and
Chronic uraemia 2% vs. 4% Y means that drug effect can NEUVONEN
be monitored clinically as 1986
clinical end point can be
determined safely
Age-related decrease 1.8% vs. 3.0% Y Pharmacokinetic changes TIULA and
in renal function but no clinical implication ELFVING
1987
Digitoxin Chronic cardiac 6% vs. 5.9% N Not significant FITCHL
failure et a1. 1983
End-stage renal 2.0% vs. 2.5% N Difference too small to LOHMAN
disease have any therapeutic and
(haemodialysis) consequence MERKUS
1978
Disopyramide MI 0.25%, vs. 0.15% Y Binding varies with drug DAVID
(at 2mgll) and protein concentration. et a1. 1983
0.53% vs. 0.32% May be clinically significant
(at 5mgll) as AGP concentration
decreases post MI
MI 0.2% vs. 0.13% Y Not indicated. CAPLIN
Pharmacokinetic change et a!. 1985
suggested. Interpretation of
total drug levels caution
Flecainide MI 39% vs. 47% Y Not indicated. Endogenous CAPLIN
compounds may lead to et a!. 1985
displacement and
decreased binding
Imipramine Chronic cardiac 19% vs. 18% N Drug binding not affected FrrcHL
failure et a1. 1983
Lignocaine Uraemia 30.7% vs. 20.8% Y Interpreting kinetic data GROSSMAN
142 J.e. McELNAY
Table 1. Continued
Drug Disease % Free or free Significant Clinical significance! References
fraction (control difference comment
vs. patient)

Kidney transplant 33.7% vs. 24.6% Y with regard to the type of et al. 1982
renal disease and the
Nephrotic syndrome 30.4 % vs. 34.2 % N protein involved in
binding may be important
NIDDM 32% vs. 30% N Not indicated (but no O'BYRNE
change expected) et al. 1988
Lorcainide Cardiac arrhythmia 26.03% vs. 24.07% N Not significant SOMANI
et al. 1984
Renal disease 26.03% vs. 29.04% N Not significant
Metociopramide Renal disease 0.6% vs. 0.59% N Not significant WEBB et al.
1986
Phenylbutazone Alcoholic hepatitis 6% vs. 13% Y Not known. BRODIE
Alcoholic cirrhosis 6% vs. 19% Y Bilirubin and and BOOBIS
hypoalbuminaemia 1978
implicated in binding
defect
Phenytoin Early pregnancy' 9.7% vs. 10.6% Y Important in drug PERUCCA
Mid pregnancy 9.7% vs. 10.9% Y monitoring interpretation. et al. 1981
Late pregnancy 9.7% vs. 12.6% Y Decreased albumin
concentration in
pregnancy implicated
Chronic cardiac = 15% vs. 15% N Not significant FITCHL
failure et al. 1983
Acute uraemia 10% vs. 25% Y Important clinically in drug TIULA and
Chronic uraemia 10% vs. 24% Y monitoring, although free NEUVONEN
concentration remains the 1986
same due to
pharmacokinetic
compensation
Age-related 10% vs. 13.5% Y Important in drug TIULA and
decrease in monitoring ELFVlNG
renal function 1987
Prednisolone Portosystemic 17.7% vs. 28.6% Y Not significant as BERGREM
shunt elimination and V d altered; et al.1983
therefore, normalising free
concentration
Nephrotic syndrome 2.26 x l(J3vs. Y Not known. Altered FREY and
4.20 x 103 M-I pharmacokinetic FREY 1984
(albumin K,) disposition, due to
2.12 x 10' vs. N change in binding
3.44 x 10' M-I
(transcortin K,)
Propranolol Chronic cardiac 14% vs. 14% N AGP binding not affected FITCHL
failure et al. 1983
Cancer, 10.8% vs. 5.5% Y Not clinically significant PAXTON
MI and IHD, but may be important when and BRIANT
infection, heart interpreting drug levels 1984
failure, COPD,
CVA, miscellaneous
Acute uraemia 10% vs. 9% N Not significant TIULA and
Chronic ureaemia 10% vs. 8.9% N Not indicated NEUVONEN
1986
Salicylate Alcoholic hepatitis 27% vs. 34% N Not known. BRODIE
Alcoholic cirrhosis 27% vs. 41% Y Bilirubin and and BooBIs
hypoalbuminaemia 1978
implicated in binding
defect
Drug Interactions at Plasma and Tissue Binding Sites 143

Table 1. Continued
Orug Disease % Free or free Significant Clinical significancel References
fraction (control difference comment
vs. patient)

Sulphadiazine Alcoholic hepatitis 46% vs. 58% Y Not known. Bilirubin BRODIE
Alcoholic cirrhosis 46% vs. 51 % N and hypoalbuminaemia and BoosTs
implicated in binding 1978
defect
Sulphisoxazole Uraemia 5.2% vs. 21.8% Not indicated. CALVO
Carbamylation of drug et a!. 1982
binding site I implicating in
defect; site II not affected
Theophylline Acute illness 54.6% vs. 69.7% Y Not significant as free ZAROWITZ
inCOPO 7.4% vs. 8.1 N concentration is not et a!. 1985
(I'glml) is not changed significantly
Tolfenamic acid Renal disease 0.08% vs. 0.17% Y Not clear. High affinity LAZNICEK
Liver disease 0.08% vs. 0.29% Y for red blood cells means and SENIUS
that these may act as 1986
reserve binding sites when
free fraction increases
Valproic acid Renal disease 8.4% vs. 20.3% Y Not clear. May lead to GUGLER
increased incidence of and
toxicity. Important in drug MUELLER
monitoring interpretation 1978
Early pregnancy' 9.4% vs. 11.5% Y Important in drug PERUCCA
Mid pregnancy 9.4% vs. 12.1 % Y monitoring. et a!. 1981
Late pregnancy 9.4% vs. 14.6% Y Oecreased albumin serum
concentration implicated in
binding defect
lOOM 6.2% vs. 7.6% Y Not significant if diabetes GAITI
is well controlled. If poorly et a!. 1987
controlled increased free
concentration may be
greater
Verapamil Arrhythmia Not indicated. MCGOWAN
Pharmacokinetic change et a!. 1982
may be implicated
Liver disease 0.099 vs. 0.16 Y Clinical end point effective GIACOMINI
to titrate dose so change et a!. 1984
not significant in therapy.
Change may be relevant
when interpreting kinetic
data and total drug
concentrations
Warfarin Chronic cardiac 1% vs.l% N No change expected FITCHL
failure et a!. 1983
Warfarin NIDOM 1.1 % VS. 1% N Not indicated (but no O'BYRNE
change expected) et a!. 1988
Zomepirac Uraemia 1.4% VS. 4% Y Decrease due to PRITCHARD
endogenous inhibitors et a!. 1983
Not clinically significant

Y, yes; N, no; MI, myocardial infarction; IHO, ischaemic heart disease; COPO, chronic obstructive pulmonary disease;
CV A, cerebrovascular accident; AGP, al-acid glycoprotein; K" drug-protein association constant; V d , apparent volume
of distribution; NIOOM, non-insulin-dependent diabetes mellitus; 100M, insulin-dependent diabetes mellitus.
, Although pregnancy is not a disease, it can give rise to changed binding; therefore it has been included in this table.
 ......
t
Is drug of interest> 90%
protein bound?

Yes
No

Does the drug have a

narrow therapeutic
index? No

Yes
-------------=: ,
What is the hepatic Low Would a transient increase No
in free drug concentration 1-1- - - - - + Clinically significant interaction
extraction ratio not likely
of the drug? be clinically relevant?

High

Is the drug
given i.v.?
Yes

Yes
....
Clinically significant (1
interaction likely.
Perform a clinical study ~
C"l
to quantify effects Fig. 7. Algorithm for determining clinical significance of potential plasma tr1
protein binding displacement interactions. (After ROLAN 1994)
~
Drug Interactions at Plasma and Tissue Binding Sites 145

I. Conclusions
Plasma protein binding displacement has been overestimated and overstated
as a mechanism of drug-drug interactions since the effects are of a transient
nature; it is only considered to be problematic with regard to adverse clinical
outcomes under specific conditions. Drug interactions, which in the past have
been attributed to plasma binding displacement, have another underlying
mechanism, usually involving decreased elimination of the drug in the liver or
kidney. ROLAN (1994) has drawn up an algorithm to help assist in determining
whether displacement interactions are likely to be clinically significant. A copy
of this algorithm is presented in Fig. 7. If a plasma protein binding interaction
is suspected, and the drug has a narrow therapeutic range, it is advisable to
monitor free (unbound) concentrations of the drug as a guide to dosage
adjustment. This is also the case in patients who have diseases which give rise
to changed binding. Tissue binding displacement interactions, since re-equili-
bration of free drug concentrations post an interaction is dependent on the
drug's half-life, rather than on redistribution, can be more problematic and
this type of interaction can be complicated by decreased elimination of the
displaced drug, as is seen in the case of the classical interaction between
digoxin and quinidine.

References
Aggeler PM, O'Reilly RA, Leong L, Kowitz PE (1967) Potentiation of anticoagulant
effect of warfarin by phenylbutazone. N Engl 1 Med 276:496-501
Aguirre C, Rodriguez-Sasiain 1M, Calvo R (1994) Decrease in penbutolol protein
binding as a consequence of treatment with some alkylating agents. Cancer
Chemother Pharmacol 34:86-88
Allen LM, Creaven PI (1974) Binding of a new antitumour agent, thalicorpine to
DNA. 1 Pharm Sci 63:74-475
Barre J, Houin G, Brunner RF, Bree F, Tillement IP (1983) Disease induced modifica-
tions of drug pharmacokinetics. Int 1 Clin Pharmacol Res 3:215-226
Baruzzi A, Contin M, Perucca E, Albani F, Riva R (1986) Altered serum protein
binding of carbamazepine in disease states associated with an increased lX:t-acid
glycoprotein concentration. Eur 1 Clin Pharmacol 31:85-89
Beck WS, Dietzel K, Geisslinger G, Engler H, Vergin H, Brune K (1990) Effects of
sodium salicylate on elimination kinetics of indomethacin and bile production in
dogs. Drug Metab Dispos 18(6):962-967
Belpaire FM, Wynant P, Van Trappen P, Dhont M, Verstraete A, Bogaert MG (1995)
Protein binding of propranolol and verapamil enantiomers in maternal and foetal
serum. Br 1 Clin PharmacoI39:190-193
Bergrem H, Ritland S, Opendal I, Bergran A (1983) Prednisolone pharmacokinetics
and protein binding in patients with portosystemic shunt. Scand 1 Gastroenterol
18:273-276
Bigger IT (1979) The quinidine-digoxin interaction. What do we know about it? N Engl
1 Med 301:779-781
Bree F, Urien S, Nguyen P, Riant P, Albengres E, Tillement JP (1990) Are-evaluation
of the HSA-piroxicam interaction. Eur 1 Drug Metab PharmacoI15(4):303-307
Brodie LMl, Boobis D (1978) The effect of chronic alcoholic ingestion and alcoholic
liver disease on binding of drugs to serum proteins. Eur 1 Clin Pharmacol13:435-
438
146 J.C. McELNAY

Calvo R, Carlos R, Erill S (1982) Effects of carbamylation of plasma proteins and

competitive displacers on drug binding in uraemia. Pharmacology 24:248-252
Caplin JL, Johnston A, Hamer J, Camm AJ (1985) The acute changes in serum binding
of disopyramide and flecainide after myocardial infarction. Eur J Clin Pharmacol
28:253-255
Chakrabarti SK (1978) Cooperativity of warfarin binding with human serum albumin
induced by free fatty acid anion. Biochem PharmacoI27:739-743
Christensen LK, Hansen JM, Kristensen M (1963) Sulphaphenazole-induced
hypoglycaemic attacks in tolbutamide-treated diabetics. Lancet 11:1298-1301
Craig W A, Evenson MA, Ramgopal V (1976) The effect of uraemia, cardiopulmonary
bypass and bacterial infection on serum protein binding. In: Benet (ed) The effect
of diseases states on drug pharmacokinetics. Am Pharm Assoc, Washington, pp
125-136
Crooks MJ, Brown KF (1974) The binding of sulphonylureas to serum albumin. J
Pharm Pharmacol 26:304-311
D'Arcy PF, McElnay JC (1982) Drug interactions involving the displacement of drugs
from plasma protein and tissue binding sites. Pharmacol Ther 17:211-220
Dasgupta A, Jaques M (1994) Reduced in vitro displacement of valproic acid from
protein binding by salicylate in uremic sera compared with normal sera. Role of
uremic compounds. Am J Clin Path 101:349-353
David BM, Ilett KF, Whitford EG, Stenhouse NS (1983) Prolonged variability in
plasma protein binding of disopyramide after acute myocardial infarction. Br J
Clin PharmacoI15:435-441
Davis DR, Yeary RA (1975) Bilirubin binding to hepatic Y and Z protein (Ligandin):
tissue bilirubin concentration in phenobarbital treated Gunn rat. Proc Soc Exp
BioI Med 148:9-13
De Sante KA, Dittert LW, Stavchansky S, Doluisio JT (1980) Influence of sulfaethidole
on the human pharmacokinetics of dicloxacillin. J Clin Pharmacol 20:535-542
Desmond PV, Roberts RK, Wood AJJ, Dunn GD, Wilkinson GR, Schenker S (1980)
Effect of heparin administration on plasma binding of benzodiazepines. Br J Clin
PharmacoI9:171-175
Doering W (1979) Quinidine-digoxin interaction: pharmacokinetics, underlying
mechanism and clinical implications. N Engl J Med 301:400-404
Doherty JE, Perkins WH, Flanigan WJ (1967) The distribution and concentration of
tritiated digoxin in human tissues. Ann Int Med 66:116-124
Ferrazzini G, Klein J, Sulh H, Chung D, Griesbrecht E, Koren G (1990) Interaction
between trimethoprim-sulfamethoxazole and methotrexate in children with leuke-
mia. J Pediatr 117(5):823-826
Fiset C, Valee F, LeBel M, Gergeron MG (1986) Protein binding of ceftriaxone:
comparison of three techniques of determination and the effect of 2-
hydroxybenzoylglycine, a drug binding inhibitor in uraemia. Ther Drug Monit
8:483-489
Fitchl B, Meister W, Schmied R (1983) Serum protein binding of drugs is not altered in
patients with severe chronic cardiac failure. Int J Clin Pharmacol Ther Toxicol
21:241-244
Frey FJ, Frey BM (1984) Altered plasma protein binding of prednisolone in patients
with the nephrotic syndrome. Am J Kidney Dis 3:339-348
Friel PN, Leal KW, Wilensky AJ (1979) Valproic acid-phenytoin interaction. Ther
Drug Monit 1:243-248
Gatti G, Grema F, Attardo-Parrinello G, Frantion P, Aguzzi F, Perucca E (1987)
Serum protein binding of phenytoin and valproic acid in insulin-dependent diabe-
tes mellitus. Ther Drug Monit 9:389-391
Giacomini KM, Giacomini JC, Blaschke TF (1980) Absence of effect of heparin on the
binding of prazosin and phenytoin to plasma proteins. Biochem Pharmacol
29:3337
Giacomini KM, Massoud N, Wong FM, Giacomini JC (1984) Decreased binding of
verapamil to plasma proteins in patients with liver disease. J Cardiovasc
Pharmacol 6:924-928
Drug Interactions at Plasma and Tissue Binding Sites 147

Grainger-Rousseau TJ, McElnay JC, Collier PS (1989) The influence of disease on

plasma protein binding of drugs. Int J Pharm 54:1-13
Grossman SH, Davis D, Kitchell BB, Shand DG, Routledge PA (1982) Diazepam and
lidocaine plasma protein binding in renal disease. Clin Pharmacol Ther 31:350-
357
Gugler R, Mueller G (1978) Plasma protein binding of valproic acid in healthy subjects
and in patients with renal disease. Br J Clin Pharmacol 5:441-446
Hager WD, Fenster P, Mayersohn M, Perrier D, Graes P, Marcus FI, Goldman S
(1979) Digoxin-quinidine interaction. N Engl J Med 300:1238-1241
Hasegawa T, Hara K, Hata S (1994) Binding of dorzolamide and its metabolite,
N-deethylated dorzolamide, to human erythrocytes in vitro. Drug Metab Disposit
22:377-382
Johannessen SI (1992) Pharmacokinetics of valproate in pregnancy: mother-foetus-
newborn. Pharm Weekbl (Sci Ed) 14:114-117
Jusko WJ, Gretch M (1976) Plasma and tissue binding of drugs in pharmacokinetics.
Drug Metab Rev 5:43-140
Kessler KM, Leech RC, Spann JF (1979) Blood collection techniques, heparin and
quinidine protein binding. Clin Pharmacol Ther 25:204-210
Kishore K, Raina A, Misra V, Jonas E (1993) Acute verapamil toxicity in a patient with
chronic toxicity: possible interaction with ceftriaxone and clindamycin. Ann
Pharmacother 27:877-880
Knott C, Hamshaw-Thomas A, Reynolds F (1982) Phenytoin-valproate interaction:
importance of saliva monitoring in epilepsy. Br Med J 284:13-16
Kober A, Ekman B, Sjoholm I (1978) Direct and indirect determination of binding
constants of drug-protein complexes with microparticles. J Pharm Sci 67:107-
109
Koch-Weser J (1979) Disopyramide. N Engl J Med 300:957-962
Koch-Weser J, Sellers EM (1976) Binding of drugs to serum albumin (first of two
parts). N Engl J Med 294:311-316
Laznicek M, Senius KEO (1986) Protein binding of tolfenamic acid in the plasma from
patients with renal and hepatic disease. Eur J Clin PharmacoI30:591-596
Lewis RJ, Trager WF, Chan KK, Breckenridge AK, Orme M, Rowland M, Schary W
(1974) Warfarin: stereochemical aspects of its metabolism and the interaction with
phenylbutazone. J Clin Invest 53:1607-1617
Liegler DG, Henderson ES, Hahn MA, Oliverio VT (1970) The effect of organic acids
on renal clearance of methotrexate in man. Clin Pharmacol Ther 10:849-857
Lindup WE (1975) Drug-albumin binding. Biochem Soc Trans 3:635-640
Lohman JJHM, Merkus FWHM (1987) Plasma protein binding of digitoxin and some
other drugs in renal disease. Pharm Weekbl (Sci Ed) 9:75-78
May T, Rambeck B (1990) Fluctuations of unbound and total phenytoin concentrations
during the day in epileptic patients on valproic acid comedication. Ther Drug
Monit 12:124-128
McAuliffe JJ, Sherwin AL, Leppik IE, Fayle SA, Pippenger CE (1977) Salivary levels
of anticonvulsants: a practical approach to drug monitoring. Neurology (Minneap)
27:409-413
McElnay JC, D'Arcy PF (1980) Displacement of albumin-bound warfarin by anti-
inflammatory agents in vitro. J Pharm PharmacoI32:709-711
McElnay JC, D'Arcy PF (1983) Protein binding displacement interactions and their
clinical importance. Drugs 25:495-513
McEinay JC, Sidahmed AM, D'Arcy PF, McQuade RD (1985) Chloroquine-digoxin
interaction. Int J Pharm 26:267-274
McGowan FX, Reiter MJ, Pritchett EI, Shand DG (1982) Verapamil plasma binding:
relationship to lX:t-acid glycoprotein and drug efficacy. Clin Pharmacol Ther
33:485-490
O'Byrne PO, O'Connor P, Feely J (1988) Plasma protein binding of lignocaine and
warfarin in non-insulin dependent diabetes mellitus. Br J Clin Pharmacol 26:
648
0ie S (1986) Drug distribution and binding. J Clin PharmacoI26:583-586
148 J.e. MCELNAY
O'Reilly RA (1971) Interaction of several coumarin compounds with human and
canine plasma albumin. Mol Pharmacol 7:209-218
O'Reilly RA (1980) Stereoselective interaction of trimethoprim-sulfamethoxazole
with the separated enantiomorphs of racemic warfarin in man. N Engl J Med
303:33-36
Paxton JW, Briant RH (1984) a)-acid glycoprotein concentrations and propranolol
binding in elderly pateints with acute illness. Br J Clin Pharmacol 18:806-810
Perrin A, Milano G, Thyss A, Cambon P, Schneider M (1990) Biochemical and phar-
macological consequences of the interaction between methotrexate and
ketoprofen in the rabbit. Br J Cancer 62:736-741
Perucca E, Hebdige S, Frigo GM, Gatti G, Lecchini S, Crema A (1980) Interaction
between phenytoin and valproic acid: plasma protein binding and metabolic ef-
fects. Clin Pharmacol Ther 28:779-789
Perucca E, Ruprah M, Richens A (1981) Altered drug binding to serum proteins in
pregnant women: therapeutic relevance. J R Soc Med 74:422-426
Piafsky KM (1980) Disease induced changes in the plasma binding of basic drugs. Clin
Pharmacokinet 5:246-262
Piafsky KM, Borga 0, Odar-Cederlof L, Johansson C, Sjovqist F (1978) Increased
plasma protein binding of propanolol and chlorpromazine mediated by disease
induced elevations of plasma a)-acid glycoprotein. N Engl J Med 299:1435-1439
Pinkard RN, Hawkins D, Farr RS (1973) The infleunce of acetylsalicylic acid on the
binding of acetrizote to human albumin. Ann New York Acad Sci 226:341-354
Pisani FD, Di Perri RG (1981) Intravenous valproate: effects on plasma and saliva
phenytoin levels. Neurology 31:467-470
Plumbridge TW, Aarons U, Brown JR (1978) Problems associated with analysis and
interpretation of small molecule/macromolecule binding data. J Pharm Pharmacol
30:69-74
Pond SM, Birkett DJ, Wade DN (1977) Mechanisms of inhibition of tolbutamide
metabolism: phenylbutzone, oxphenbutazone, sulphaphenazole. Clin Pharmacol
Ther 22:573-579
Powell-Jackson PR (1977) Interaction between azapropazone and warfarin. Br Med J
1:1193-1194
Pritchard JF, O'Neill PJ, Affrime MB, Lowenthal DT (1983) Influence of uraemia,
haemodialysis and nonesterified fatty acids on zomepirac plasma protein binding.
Clin Pharmacol Ther 34:681-688
Reiffel JA, Leahey EB, Drusin RE, Heissenbuttel RH, Lovejoy W, Bigger JT (1979) A
previously unrecognised drug interaction between quinidine and digoxin. Clin
Cardiol 2:40-42
Rolan PE (1994) Plasma protein binding displacement interactions - why are they still
regarded as clinically important? Br J Clin PharmacoI37:125-128
Rothschild MA, Oratz M, Schreiber SS (1973) Albumin metabolisl)l. Gastroenterol
64:324-337
Rowland M (1980) Plasma protein binding and therapeutic drug monitoring. Ther
Drug Monit 2:29-37
Rowland M, Tozer TN (1995) In: Clinical pharmacokinetics, concepts and applications.
Williams and Wilkins, Philadelphia, pp 270-289
Salazar-Bookaman MM (1994) Relevance of drug-melanin interactions to ocular phar-
macology and toxicology. J Ocul PharmacoI1O:217-239
Scatchard G (1949) The attractions of protein for small molecules and ions. Ann NY
Acad Sci 51:660-672
Sellers EM, Koch-Weser J (1970) Potentiation of warfarin induced hypopro-
thrombinemia by chloral hydrate. N Engl J Med 283:827-831
Shaklai N, Garlick RL, Bunn HF (1984) Nonenzymatic glycosylation of human serum
albumin alters its conformation and function. J BioI Chern 259:3812-3817
Somani P, Simon V, Gupta RK, King P, Shapiro RS, Stockard H (1984) Lorainide
kinetics and protein binding in patients with end-stage renal disease. Int J Clin
Pharmacol Ther ToxicoI223:121-125
Drug Interactions at Plasma and Tissue Binding Sites 149

Stewart CF, Fleming RA, Germain BF, Seleznick MJ, Evans WE (1991) Aspirin alters
methotrexate disposition in rheumatoid arthritis patients. Arthritis Rheum
34:1514-1520
Tillement JP (1980) Plasma binding of drugs. Pharm Int 1:64-65
Tillement JP, Lhoste F, Gidicelli JF (1978) Disease and drug protein binding. Clin
Pharmacokinet 3:144-154
Tiula E, Elfving S (1987) Serum protein binding of phenytoin, diazepam and
propranolol in age-related decrease in renal function. Ann Clin Res 129:163-169
Tiula E, Neuvonen PJ (1986) Effect of total drug concentration on the free fraction in
uraemic sera. Ther Drug Monit 8:27-31
Toon S, Low L, Gibaldi M, Trager WF, O'Reilly RA, Motley ChH (1986) The war-
farin-sulphinpyrazone interaction: stereochemical considerations. Clin Pharmacol
Ther 39:15-24
Trnavska Z, Krejocova H, Tkaczykovam, Salcmanova Z, Elis J (1991) Pharmacokinet-
ics of lamotrigine (Lamictal) in plasma and saliva. Eur J Drug Metab Pharmacokin
3:211-215
Webb D, Buss DC, Fifield R, Bateman DN, Routledge PA (1986) The plasma protein
binding of metoclopramide in health and renal disease. Br J Clin Pharmacol
21:334-336
Wood M, Shand DG, Wood AJ (1979a) Altered drug binding due to use of indwelling
heparinized cannulas (heparin lock) for sampling. Clin Pharmacol Ther 25:103-
107
Wood M, Shand DG, Wood AJ (1979b) Propranolol binding in plasma during cardio-
pulmonary bypass. Anesthesiol. 51:512-516
Zarowitz B, Shlom J, Eichenhorn MS, Popovich J (1985) Alternations in theophylline
protein binding in acutely ill patients and COPD. Chest 87:766-769
CHAPTER 5
Drug Interactions
and Drug-Metabolising Enzymes
P.F. D'ARCY

A. General Introduction
Induction and inhibition of cytochrome P450 enzymes are mechanisms that
underly some of the more serious drug-drug interactions; it is therefore perti-
nent that the structure, distribution and essential roles of cytochromes P450
are reviewed.
Cytochrome P450 is not a single species of protein; the system is actually
a collection of isoenzymes, all of which possess an iron atom in a porphyrin
complex. They catalyse different types of oxidation reactions and under cer-
tain circumstances may catalyse other types of reaction such as reduction
(TIMBRELL 1993).
Since the early experiments of CONNEY et al. (1956; CONNEY 1967) it has
often been observed that the rate of oxidative metabolism of various sub-
strates can differ markedly depending on the age, sex, species or the extent of
exposure of the animal to different inducing agents. A number of different
investigators have directed studies to evaluate the number and types of cyto-
chrome P450s that exist in a single organ. In particular they questioned
whether specific types of reactions catalysed by the microsomal electron trans-
port system required specific cytochrome P450, or whether many cytochrome
P450s have a rather broad substrate specificity differing only in the rate of
catalysis of each reaction.
SCHENKMAN (1993) has well documented the beginning of this interest and
study into the cytochrome P450 monooxygenase. It commenced with the
compilation of knowledge of the metabolism of xenobiotics by WILLIAMS
(1959), who described the many different metabolites produced from
xenobiotics in vivo by animals. From the late 1940s into the 1960s knowledge
was advanced by BRODIE et al. (1955), who was one of the first to start in vitro
biochemical studies on the metabolism of xenobiotics by oxidative enzymes.
Work from MILLER'S laboratory augmented these studies (MULLER and
MILLER 1953) and it soon became clear that liver micro somes were the source
of NADPH-dependent, oxidative enzymes capable of metabolising a number
of xenobiotics.
Since these early studies in the 1950s more than 800 different xenobiotics,
many of which are therapeutic drugs, have been shown to be substrates
for liver microsomal oxidative enzymes. The major types of oxidation
152 P.F. D'ARCY

reaction catalysed by the cytochrome P450 system can be subdivided

into: oxidation or hydroxylation (e.g., many drugs including paroxetine),
deamination (e.g., amphetamine), de alkylation (e.g., morphine), sulphoxi-
dation (e.g., chlorpromazine, paroxetine), desulphuration (e.g., thiopentone),
dehalogenation (e.g., halogenated anaesthetics) and glucuronidation (e.g.,
paroxetine ).
Certain oxidative reactions of xenobiotics are also catalysed by enzymes
other than the cytochrome P450 monooxygenase system, for example the
microsomal FAD-containing monooxygenase (ZEIGLER 1984, 1985; DAMANI
1988). This is responsible for the N-oxidation of tertiary amines such as
dimethylaniline and trimethylamine. The enzyme requires NADPH and
oxygen; the substrate specificity also includes secondary amines and
sulphides, thioethers and thiocarbamates, and organophosphates (TIMBRELL
1993). Alcohols may be oxidised by alcohol dehydrogenase; xanthine
oxidase catalyses the oxidation of nitrogen heterocyclics such as the purine
hypoxanthine. Some amines such as tyramine are substrates for monoamine
oxidases; diamines such as putrescine are metabolised by the soluble enzyme
diamine oxidase. The peroxidases are also involved in the oxidation of
xenobiotics, the most important being prostaglandin synthase, which is known
to catalyse the oxidation of p-phenetidine, a metabolite of phenacetin, a
process that may be involved in the nephrotoxicity of the drug (TIMBRELL
1993).
Extensive work by many investigators including AXELROD (1955), POSNER
et al. (1961), SLADEK and MANNERING (1966), ALVARES et al. (1967) and Lv and
COON (1968) isolated and purified a number of cytochrome P450 genes from
animals and at the current time some 154 cytochrome P450 genes have been
characterised from humans, animals, insects and plant species.
During the last 10 years or so, a vast amount of work has been
done on the identification and characterisation of human xenobiotic-
metabolising cytochrome P450s. Most studies have been done on hepatic and
pulmonary enzymes due to the availability of tissues from tissue donor
programmes.
BLACK (1993) has reviewed the cytochrome P450 structure and function
and GONZALEZ (1993) cytochrome P450 in humans. Mammalian P450 en-
zymes are tightly bound in both microsomes (endoplasmic reticulum) and
mitochondria and have been purified by means of detergent solubilisation.
Purified detergent-free preparations exhibit a highly amphiphilic character
and exist as micellar aggregates of approximately six protomers (DEAN and
GRAY 1982). The enzymes are discrete gene products of about 57000 molecu-
lar weight and contain one equivalent of b-type heme per polypeptide.
The most abundantly expressed cytochrome P450s in human liver are the
CYP3A subfamily. A brief description on this subfamily will serve as an
example of the whole genus. Other subfamilies in human tissues comprise
CYP1A, CYP2B, CYP2C, CYP2D, CYP2E, and CYP2F; the locus, function
and activities of these are shown in Table 1.
Drug Interactions and Drug-Metabolising Enzymes 153

Table 1. Cytochromes P450 in human. (Based on GONZALEZ 1993)a

Subfamily Members Chromosome Established functions/activities

(locus)
Metabolises Induced by

CYP1A1 CYP1A1 15q22-qter Aflatoxin B J Ketocoazole

CYP1A2 Arylamine and its Omeprazole
promutagens Polycyclic
Acetaminophen aromatic
Benzo[ a]pyrene hydrocarbons
Caffeine Tobacco smoke
Food mutagens
Phenacetin
Warfarin
CYP2A CYP2A6 19q12-13.2 Aflatoxin B J
CYP2A7 N-Nitrosodiethylamine
Warfarin
CYP2B CY2B6 19 Warfarin Phenobarbitone
CY2B7
CYP2C CY2C9 lOq24.1-24.3 Mephentoin
CYP2C10 Tolbutamide
CYP2C17 Warfarin
CYP2C18
CYP2C19
CYP2D CYP2D6 22ql1-2qter Debrisoquine Tobacco smoke
CYP2D7P Sparteine
CYP2D8P Nitrosamine
Carcinogen (NNK)
CYP2E CYP2E1 10 Acetone Ethanol
Caffeine
Chlorzoxazone
Paracetamol
Solvents
N -Nitrosodimethylamine
CYP2F CYP2F1 19 Naphthylamine
CYP3A CYP3A3 7q21-22.1 Aflatoxins Dexamethasone
CYP3A4 Benzo[ a]pyrrene Steroids
CYP3A5 Cyclosporin
CYP3A7 Erythromycin
17 -Ethynyloestradiol
Lidocaine
Midazolam
Nifedipine
Quinidine
Warfarin

aOther cytochrome P450s have been detected in animals; multiple gene conversions
have occurred among some animal CYP genes, rendering them quite dissimilar to their
counterparts in humans.
154 P.F. D'ARCY

CYP3A cytochrome P450s account for up to 60% of total cytochrome

P450 in some liver specimens (SHIMADA and GUENGERICH 1989; GUENGERICH
and KIM 1990). A CYP3A protein has been isolated from human livers on the
basis of its ability to oxidise the calcium channel blocking drug nifedipine
(GUENGERICH et al. 1986). Interestingly, a cytochrome P450, designated
CYP3A7, that is absent in adults has been found in foetal human liver. Al-
though this is the most abundantly expressed cytochrome P450 in the foetus,
its function is unknown.
There is a large degree of interindividual variability in the expression of
CYP3A proteins, this could be due to the presence or absence of inducers or
due to a genetic polymorphism. CYP3A proteins are also abundantly ex-
pressed in extrahepatic tissues, especially the gut (WATKINS et al. 1987;
DEWAZIERS et al. 1990). This is believed to have important implications for the
metabolism of orally administered erythromycin, and other drugs that serve as
a substrate for this enzyme. Immunocorrelation studies have established that
CYP3A4 and probably CYP3A3 metabolise numerous drugs and toxins in-
cluding afiatoxins, benzo[ a]pyrene, cyclosporin, erythromycin, lidocaine, 17-
ethynyloestradiol, midazolam, quinidine and warfarin.
This subfamily is therefore important not only for drug clearance but also
for carcinogen activation, properties shared by many of the other subfamilies
of cytochrome P450s. The structure of human CYP3A genes has not yet been
determined but the locus has been found on chromosome 7 (7q21-q22.1)
(BROOKS et al. 1988).

B. Extrahepatic Microsomal Forms of Cytochrome P450

Cytochrome P450 has been shown to exist in many extrahepatic tissues in
human and animals, including the gastrointestinal tract, lung and brain.
Hydroxylation of benzo[a]pyrene has been induced in the oesophagus,
forestomach, duodenum, small intestine, caecum and colon of animals
treated with 1,2-benzanthrene, a well-known inducer of cytochrome P450
(WATTENBERG et al. 1962; WATTENBERG and LEON 1962). The small intestine
has been demonstrated to catalyse the de alkylation of phenacetin (PANTUCK et
al. 1975) as well as 7-ethoxycoumarin. The presence of an active cytochrome
P450-dependent mixed function oxidase system in tumour cell microsomes
illustrates the potential for both drug metabolism and carcinogen activation
(STROBEL et al. 1993).
Cytochrome P450 content and activity are quite low in the lung when
compared with those of the liver. Much of the investigative work has been
done using rabbit lung microsomes (see, for example, PHILPOT et al. 1975;
WILLIAMS et al. 1984; MATSUBARA et al. 1987) since this species has higher
cytochrome P450 content and activities than other animal species and man
(see, for example, PHILPOT et al. 1975; McMANUS et al. 1980). The lungs
contain several cytochrome P450 isoenzymes with different substrate specifici-
ties (see review by PHILPOT and WOLF 1981).
Drug Interactions and Drug-Metabolising Enzymes 155

The presence of cytochrome P450s in the lung is of obvious importance

since this tissue is the portal of entry of various airborne contaminants and
potential carcinogens and, in so far as therapy is concerned, it is a site for
absorption of a variety of aerosolised inhalants and a site for drugs reaching
the lung via the circulatory system (ARINc; 1993).
Microsomal and mitochondrial forms of cytochrome P450 have been
found in the brain although in low concentrations (see review by WARNER and
GUSTAFSSON 1993). Most of the cytochrome P450 in the brain remains
uncharacterised. The brain is not very responsive to well-known inducers of
hepatic P450 such as barbiturates and polycyclic aromatic hydrocarbons
(SRIVASTAVA and SETH 1983). A notable exception to this resistance to induc-
tion is the long-term administration of phenytoin (VOLK et al. 1988).
Brain cytochrome P450 is hormonally regulated and there is a five to
tenfold increase of P450 content in the hypothalamic preoptic area and
the olfactory lobes during pregnancy, lactation and in response to
dihydrotestosterone treatment during pregnancy and lactation. Induction
occurs in both the microsomal and mitochondrial fractions (WARNER et al.
1989).
Novel physiological functions of brain cytochrome P450 are thought to
include:
1. Regulation of the intracellular levels of cholesterol (GUPTA et al. 1986;
RENNERT et al. 1990).
2. Signal transduction - met abo Ii ties of arachidonic acid, formed through
action of cytochrome P450, are thought to be the intracellular signals in-
volved in the release of peptide hormones from the hypothalamus and
pituitary (SNYDER et al. 1983; CASHMAN et al. 1987).
3. Regulation of vascular tone via P450-dependent metabolites of arachidonic
acid (SACERDOTI et al. 1988; MURPHY et al. 1988).
4. Activation or inactivation of neurosteroids, e.g., dehydroepiandrosterone
synthesised in the brain through the activity of cytochrome P450scc
is believed to have a role in neuronal function (LE GOASCOGNE et al.
1987).
The role of brain cytochrome P450 in central nervous system toxicity through
the activation of procarcinogens, the formation of cytotoxic metabolites from
industrial solvents and the generation of reactive oxygen radicals during the
catalytic cycle needs to be clarified. Aromatic hydrocarbons, N-nitroso com-
pounds, triazines, hydrazines and common industrial solvents can produce
gliomas in animals (MALTONY et al. 1982). Some of the more common solvents
in paints are substrates for cytochrome P450 and exposure to industrial sol-
vents is known to produce neurotoxicity in man (MALLOV 1976; ELOFSSON et al.
1980; MAIZLISH et al. 1985). It is also a distinct possibility that there is in situ
formation of carcinogens from procarcinogens and that some of the toxicity of
these solvents is due to P450-driven metabolism within the CNS (WARNER and
GUSTAFSSON 1993).
156 P.F. D'ARCY

c. Genetic Polymorphism
There are wide variations in cytochrome P450 activity in both humans and
animals. DALY and IDLE (1993) have suggested, in their comprehensive review
on animal and human cytochrome P450 polymorphisms, that this variation
may be due to either a genetic polymorphism which results in either lack of
protein synthesis or synthesis of a protein with reduced enzyme activity, or to
a variation in the level of cytochrom P450 gene expression due to regulation by
factors such as hormones or xenobiotics.
The inability of up to 12% of individuals from a variety of racial
groups to metabolise the antihypertensive agent debrisoquine to 4-
hydroxy-debrisoquine has been known since the 1970s (MAHGOUB et al. 1977).
Pedigree studies showed that the poor metaboliser phenotype was inherited
in an autosomal recessive manner (EVANS et al. 1980). Similar findings
have been reported for the antiarrhythmic and oxytocic agent sparteine
(EICHELBAUM et al. 1979), and other compounds including phenformin (OATES
et al. 1982) and bufuralol (DAYER et al. 1982). Further studies indicated that
a single enzyme was responsible for the metabolism of all these drugs
(BERTILSSON et al. 1980). At least 25 other compounds including beta-blockers,
antihypertensives, antiarrhythmics and antidepressants are known to be
metabolised by debrisoquine hydroxylase (see review by BROSEN and GRAM
1989).
Cloning and sequencing of the CYP2D genes has assisted identification of
the mutations which give rise to the debrisoquine polymorphism in humans,
and a number of variant alleles associated with the poor phenotype have been
identified. Debrisoquine phenotype may have important consequences with
regard to both response to drug therapy and exposure to xenobiotics. For
example, the use of the tricyclic antidepressant nortriptyline, the neuroleptic
perphenzazine, or the antiarrhythmics propafenone and flecainaide may cause
problems in both rapid extensive metabolisers (low and inadequate serum
concentrations) and in poor metabolisers (toxic serum concentrations)
(BROSEN and GRAM 1989). Fortunately, however, the majority of patients
exhibit a normal metabolism of these agents and gain maximum therapeutic
benefit with minimal toxicity.
Some interesting studies on a possible association between susceptibility
to lung cancer and debrisoquine metabolic phenotype have been carried out
by AYESH et al. (1984). They found that only 1.6% of a group of bronchial
carcinoma patients were poor metabolisers of debrisoquine compared
with 9.0% of a group of smoking controls. The molecular basis of this
possible association between the poor metaboliser phenotype and decreased
lung cancer susceptibility has been suggested by CRESPI et al. (1991) to relate
to a reduced ability to activate tobacco smoke-derived procarcinogens.
An association between the poor metaboliser phenotype and susceptibil-
ity to early-onset Parkinson's disease has been suggested by BARBEAU et al.
Drug Interactions and Drug-Metabolising Enzymes 157

(1985) as possibly due to environmental exposure to pesticides similar to MPP+

which can produce a parkinsonian syndrome in man (DAVIS et al. 1979).
However, subsequent studies have failed to confirm these suggestions.
Debrisoquine was an early example used to demonstrate polymorphism;
other drugs including the anticonvulsant mephenytoin, the hypoglycaemic
tolbutamide, the antihypetensive nifedipine, and antipyrine and caffeine have
all been shown to exhibit metabolic polymorphism in humans. Apparently
tolbutamide polymorphism is distinct from the debrisoquine polymorphism
but is allied to that of phenytoin. Nifedipine metabolic polymorphism is cur-
rently somewhat debatable; the cytochrome P450 responsible for the majority
of nifedipine metabolism is the glucocorticoid-inducible CYP3A4, although
10%-20% of human livers contain an additional nifedipine-metabolising
cytochrome P450, CYP3A5. Thus variability in nifedipine metabolism may
be the result of several factors including glucocorticoid levels and CYP3A5
expression.
Similarly, there is a possible polymorphism in the metabolism of
coumarins. Cytochrome P450 CYP2A6 shows coumarin-7-hydroxylase activ-
ity and a cDNA variant ofthis enzyme CYP2A6v, which lacks enzyme activity,
may give rise to polymorphism in the metabolism of coumarin. CYP1A2 is
the enzyme responsible for the oxidative metabolism of caffeine and is
polymorphic, with approximately 40% of an American population showing
rapid metabolism of this agent [see review by DALY and IDLE (1993) and
references therein].

D. Age and Disease and Cytochrome P450

Age-related changes in pharmacokinetic behaviour, specifically of drug
metabolism, have often been attributed to changes in the cytochrome P450
based enzymes. However, studies in humans have suggested that the changes
observed in clearance of oxidatively metabolised drugs are more likely due to
physiological changes in the elderly rather than to changes in their cytochrome
P450 systems. There are few, if any, detectable changes related to ageing
in cytochrome P450 protein, mRNA or activities associated with ageing
(SCHMUCKER et al. 1990).
There have been relatively few studies on the influence of disease states on
metabolism of xenobiotics. As might be expected, different diseases of the
liver have been investigated but have been found to influence metabolism
differently (HOYUMPA and SCHENKER 1982). Indeed, some metabolic pathways
do not seem to be influenced by liver damage at all, for example the
glucuronidation of paracetamol, morphine and oxazepam was not affected by
liver cirrhosis in man although the oxidation of barbiturates, antipyrine, and
methadone and the conjugation of salicylates with glycine were all depressed
by cirrhosis (TIMBRELL 1993). Clinical influenza is known to affect drug-
158 P.F. D'ARCY

metabolising enzymes as also is influenza vaccine and possibly other

immunisation procedures. It is likely that the latter are in response to the
production of interferon (see D'ARCY 1984a; GRIFFIN et al. 1988).
Study of these aspects of drug metabolism in the elderly is important since
about 20% of the population of Western-developed countries is over the age
of 65 years and the elderly are one of the fastest-growing segments of the
community. This has importance in the pharmacotherapy since this segment is
also the heaviest consumer of medicines. It is a proven fact that the extensive
use of medication increases the possibility of adverse reactions to therapy and
it seems likely that this predisposition in the elderly is related to pharmacoki-
netic or pharmacodynamic changes and not to senescent changes in the induc-
ibility or nature of their hepatic or extrahepatic P450-driven drug metabolism
(see review by BIRNBAUM 1993).

E. Clinical Importance of Enzyme Induction or Inhibition

Many of the major interactions between drugs are due to hepatic P450
enzymes being affected by the previous administration of other drugs. Some
drugs act as potent enzyme inducers whilst others inhibit the enzyme systems.

I. Enzyme Inducers
Enzyme induction has been defined by GELEHRTER (1979) as an adaptive
increase in the number of molecules of a specific enzyme, secondary either to
an increase in its rate of synthesis or to a decrease in its rate of degradation. It
is believed that high substrate concentration, and not hormonal regulation, is
responsible for increased enzyme synthesis. If the demands on an enzyme
system exceed its maximal functional capacity, adaptation should entail a
parallel increase in enzymes and their containing structure.
Cytochrome P450 and its associated electron transport carriers are mem-
brane-bound proteins in mammalian tissue; thus the induction of hepatic
microsomal enzymes by phenobarbitone and other drugs has been shown
conclusively to stimulate an increase in the smooth endoplastic reticulum,
synthesis of enzyme protein, phospholipid and haem, DNA-dependent RNA
synthesis, and the synthesis of specific mRNAs (REMMER and MERKER 1963;
PARKE 1975; LINDELL et al. 1977). A further study of the effect of
phenobarbitone on rat liver also showed that there were no significant alter-
ations of the activities of any of the DNA-dependent RNA polymerases,
which suggested that the primary stimulus of enzyme induction was more
likely to be on post-transcriptional events such as enhanced stabilisation or
decreased turnover of hepatic RNA, than on the transcription of new RNA
(LINDELL et al. 1977).
Increased protein synthesis also appears to occur in the induction of the
hepatic microsomal mixed-function oxidases following treatment of animals
Drug Interactions and Drug-Metabolising Enzymes 159

with the carcinogenic polycyclic hydrocarbons and, as with phenobarbitone,

microsomal ribonuclease is inhibited. There are, however, a number of differ-
ences between these two inducing agents and these include an absence of
marked hypertrophy, no marked increase in phospholipid synthesis, an in-
crease of cytochrome P448 and not P450 and a change in substrate specificity
(PARKE 1979).
High demands for protein and haem synthesis and the synthesis of choline
for phosphatidylcholine make corresponding demands on dietary folate. The
prolonged administration of the enzyme inducer phenobarbitone to rats on a
low-folate diet has led to inhibition of the expected P450 microsomal induction
and to increased folate deficiency. Rats treated daily with phenytoin plus
phenobarbitone show evidence of folate deficiency.
The rate of metabolism of alcohol in alcoholics who have recently been
drinking can be more than twice that in abstinent subjects. This aspect of
induced metabolism is non-specific and also extends to many drugs. For ex-
ample, the serum level of the anticonvulsant drug phenytoin falls more rapidly
in alcoholics than it does in abstinent controls. This reduces the duration of
anticonvulsant action and there is a very real danger of precipitating seizures
if large amounts of alcohol are taken. Likewise the oral hypoglycaemic agent
tolbutamide and the anticoagulant warfarin have a shorter half-life in the
blood of alcoholics than in that of abstinent controls due to alcohol having an
inductive effect of liver microsomal enzymes.
Alcohol is often the "silent" cause of drugs failing to achieve their full
therapeutic effect. Interactions between alcohol and medication have been
reviewed by D'ARCY and MERKUS (1981).
Exposure to tobacco smoke, in both active and passive smoking, will
induce liver enzymes. A single injection of 3,4-benzpyrene, a common compo-
nent of tobacco smoke, can induce rats to synthesise an enzyme system in the
liver micro somes and non-hepatic tissues that is fully capable of hydroxylating
3,4-benzpyrene to non-carcinogenic products (CONNEY et al. 1957). Plasma
levels and clinical effects of the analgesics pentazocine and propoxyphene are
reduced in smokers. Interactions between tobacco smoking and drugs have
been reviewed by D'ARCY (1984b).
Environmental pollutants may also affect P450 activity; thus chlorinated
hydrocarbon pesticides are potent enzyme inducers and can exert long-lasting
effects due to their storage in body fat; for example, KOLMODIN et al. (1969)
showed accelerated metabolism of phenazone in farm workers exposed to
DDT and gamma benzene hexachloride (lindane).
Theoretically, the same enzyme systems may be used to reduce the serum
level of pesticides; the use of phenytoin and phenobarbitone anticonvulsants
has been suggested as a means of lowering body pesticide levels when these
become excessive. Certainly it has been shown that the DDT level in milk
from dairy cows was reduced when they were treated with phenobarbitone.
Some antibiotics are enzyme inducers and, if used at constant dosage for
long periods of time, will stimulate their own hepatic metabolism, which may
160 P.P. D'ARCY

then become the prime cause of their own inadequate levels in the blood. This
problem has particular relevance in, for example, the long-term treatment of
tuberculosis. ACOCELLA et al. (1972) have shown that rifampicin administered
constantly at 600mg/day for 7 days gave significantly lower blood levels on the
7th day than it did on the 1st day of treatment. Development of tolerance to
drug regimens is certainly caused, at least in part, by enzyme induction.
Because so many therapeutic drugs have been shown to be implicated in
cytochrome P450 induction, it is obviously impossible to describe or discuss
the effects of their interactions even concisely in a chapter of this type. As a
compromise therefore some examples of common P450-inducing agents are
given in Table 2, while some recent reports of serious drug-drug interactions
due to enzyme induction are shown as examples in Table 3.
In surveying the literature it soon becomes evident that reports of inter-
actions due to enzyme inhibition are far more numerous than those caused by
enzyme induction. This is not really surprising since the former are more
readily observed in the clinic and they evoke therapeutic overdosage or toxic-
ity, while the latter cause a reduced therapeutic effect of medication which
may not immediately be noticed or reported.

II. Enzyme Inhibitors

There are many examples of drugs slowing or inhibiting the metabolic degra-
dation of other drugs and thus causing an increase in both their duration and
intensity of pharmacological actions. Drugs causing P450 enzyme inhibition
include chloramphenicol, the H 2-receptor blocker, cimetidine, monoamine
oxidase inhibitors (MAOIs) (e.g., iproniazid, nialamide), p-aminosalicylic acid

Table 2. Some examples of compounds and drugs acting as enzyme inducers

Alcohol
Chlorinated hydrocarbon pesticides
Polycyclic hydrocarbons (e.g., dioxin) and other environmental pollutants
Tobacco smoking
Many therapeutic drugs, including:
Barbiturates
Carbamazepine
Carbenoxolone
Ciprofioxacin
Clofibrate
Glucagon
Glucocorticoids (e.g., dexamethasone)
Isoniazid
Phenybutazone
Phenytoin
Primidone
Rifampicin
Theophylline
Drug Interactions and Drug-Metabolising Enzymes 161

Table 3. Examples of serious drug-drug interactions due to enzyme induction

(illustrative only, not comprehensive)

Drug combination Sequelae

Ciprofioxacin/phenytoin Lowered phenytoin plasma levIes; danger of
breakthrough seizures in epileptics (DILLARD et al.
1992)
Combined oral Increased clearance of the anticoagulant in seven
contraceptive/phenprocoumon patients reported; bleeding may occur if oral
contraceptives are withdrawn (MONIG et al. 1990)
Phenytoin/theophylline Both drugs show a reduced plasma concentration;
poor seizure control and poor control of asthma
may result (ADEBAYO 1988)
Phenytoin/verapamil Subnormal and subtherapeutic plasma
concentrations of verapamil (WOODCOCK et al.
1991)
Rifampicin/cyclosporin Reduced cyclosporin plasma concentrations;
kidney graft rejected (LANGHOFF and MADSEN
1983)
Rifampicin/dapsone Subtherapeutic plasma concentrations of dapsone
in HIV-infected patients suffering from
Pneumocystis carinii pneumonia (HOROWITZ et al.
1992)
Rifampicin/disopyramide Effective plasma concentrations of the
antiarrhythmic were difficult to achieve in spite of
dosage increase; on stoppage of rifampicin it took
3-5 days for inductive effect to disappear (STAUM
1990)
Rifampicin/oral Reduction of oral contraceptive efficacy;
contraceptives breakthrough bleeding and danger of unplanned
pregnancy (D'ARCY 1986; ANGLE et al. 1991)
Rifampicin/propafenone Reduced propafenone plasma concentrations
and re-emergence of ventricular arrhythmias in a
previously stabilised patient (CASTEL et al. 1990)

(PAS), pheniprazine, the two narcotics pethidine and morphine, and the anti-
bacterial 4-quinolones. Some other agents, for example, chlorcyclizine,
glutethimide and phenylbutazone, can inhibit or induce drug metabolism
by enzymes depending upon whether they are administered acutely or chroni-
cally. Inhibition generally requires only a single dose of a compound rather
than repeated doses as does induction; the environmental impact of inhibition
is probably less than that of induction although inhibition may also be
relevant in the work place, for example, workers exposed to the solvent
dimethylformamide are less likely to suffer alcohol-induced flushes than those
not exposed, possibly due to the inhibition of alcohol metabolism (TIMBRELL
1993).
162 P.F. D'ARCY

Enzyme inhibition is especially important in relation to the therapeutic

use of MAOIs and tricyclic depressants, the coumarin anticoagulants, the
oral hypoglycaemics, the xanthine oxidase inhibitors (e.g., azathioprine,
mercaptopurine) and the immunosuppressants (e.g., cyclosporin). Many of
these drug-drug interactions can have serious, even life-threatening, sequelae.
Table 4 presents information on known inhibitors of P450 enzymes while
Table 5 gives examples of serious drug-drug interactions caused by this
mechanism.
Because of the clinical importance of many drug-drug interactions involv-
ing P450 inhibition, it is important that the mechanisms of enzymatic inhibi-
tion by drugs be understood. Unfortunately, the nature of the procedures
involving inhibition of P450 enzymes is not as well understood as the mecha-
nisms of enzyme induction.
TIMBRELL (1993) in his comprehensive review of the biotransformation
of xenobiotics, has explained that there are many different types of inhibitor
of the microsomal monooxygenase system. There are reversible inhibitors
which appear to bind as substrates and are competitive inhibitors, such as

Table 4. Some examples of compounds and drugs acting as enzyme inhibitors

Cobalt chloride
Industrial solvents (e.g., carbon tetrachloride, dimethylformamide)
Organophosphorus compounds
Many therapeutic drugs including:
Allopurinol
Amiodarone
Chloramphenicol
Cimetidine
Corticosteroids
Cyclophosphamide
Danazol
Dextropropoxyphene
Dicoumarol
Erythromycin
Felbamate
Fluconazole
Glibenclamide
Infiuenza vaccine
Iproniazid
Itraconazole
Ketoconazole
MAOI antidepressants
Moricizine
Nicoumalone
4-Quinolone and fiuoroquinolone antibacterials (e.g., ciprofioxacin, enoxacin,
lomefioxacin, norfioxacin, ofioxacin, perfioxacin)
Selegiline
Serotonin re-uptake inhibitor antidepressants (SRIs)
Tamioxifen
Triacetyloleandomycin
Drug Interactions and Drug-Metabolising Enzymes 163

Table 5. Serious drug-drug interactions due to enzyme inhibition (illustrative only, not
comprehensive)

Drug combination Sequelae

Allopurinollcyclosporin Cyclosporin plasma levels grossly increased; danger of

nephrotoxicity (GORRIE et al. 1994)
Amiodarone/theophylline Theophylline serum concentrations increased;
theophylline intoxication reported (SOTO et al. 1991)
Cyclosporin/nifedipine Competition for metabolism by cytochrome P450pcn
led to severe flushing and other signs of nifepidine
toxicity in psoriatic patients (McFADDEN et al. 1989)
Danazollwarfarin Anticoagulant effect potentiated; bleeding
complications possible; other mechanisms may be
involved (MEEKS et al. 1992; BOOTH 1993)
Erythromycin/tacrolimus Dramatic rise in plasma concentrations of the
immunosuppressant, tacrolimus; danger of CNS
toxicity or nephrotoxicity (SHAEFFER et al. 1994)
Felbamate/warfarin Anticoagulation enhanced; 50% reduction of warfarin
dosage required (TISDEL et al. 1994)
Fluconazole/nortriptyline Steady-state serum concentration of nortriptyline
increased when fluconazole was initiated and toxicity
was evident (GANNON and ANDERSON 1992)
Fluconazole/phenytoin Blood phenytoin concentration raised; danger of
serious phenytoin toxicity (CADLE et al. 1994)
Fluconazole/warfarin Over anticoagulation in a stabilised patient; bleeding
episodes reported (SEATON et al. 1991)
Fluoroquinolones/warfarin Anticoagulant effects enhanced by fluoroquinolones;
danger of bleeding (CONSUMERS' ASSOCIATION 1993)
Glibenclamide/warfarin Anticoagulation potentiated; bruising and soft tissue
bleeding with large haematomas reported (JASSAL
1991)
Ketoconazole or Plasma concentrations of the antihistamine astemizole
itraconazollastemizole or increased greatly; serious cardiac arrhythmias reported
terfenadine (AHMAD 1992; NOTICEBOARD 1992)
Moricizine/warfarin Anticoagulant effects enhanced; danger of bleeding.
Other mechanisms may also apply (SERPA et al. 1992)
Nicardipine and other Increased cyclosporin blood concentrations; danger of
Ca+-channel blockersl nephrotoxicity; other mechanisms may be involved
cyclosporin (TnA et al. 1989)
Proguanillwarfarin Anticoagulant effects enhanced; severe haematuria
(ARMSTRONG et al. 1991)
Selegiline/pethidine Enhanced toxicity of pethidine; life-threatening
interaction reported (ZORNBERG et al. 1991)
Tamoxifen/warfarin Grossly elevated anticoagulation; bleeding disorders
reported; mechanisms other than enzyme inhibition
may also apply (LODWICK et al. 1987; COMMITTEE ON
SAFETY OF MEDICINES 1989; TENNI et al. 1989)
164 P.F. D'ARCY

dichlorobiphenyl, which inhibits the O-demethylation of p-nitroanisole. There

are inhibitors which are metabolised to compounds which bind strongly to the
active site of the enzyme, such as piperonyl butoxide, which probably acts by
forming an inactive complex with cytochrome P450, which becomes irrevers-
ibly inhibited. Other inhibitors, such as carbon tetrachloride, cyclophospha-
mide, carbon disulphide and allylisopropylacetamide, destroy cytochrome
P450. Finally, there are those which interfere with the synthesis of the enzyme
such as cobalt chloride, which interferes with the synthesis of the haem portion
of the enzyme.
Other enzymes which are also involved in xenobiotic metabolism may also
be inhibited, for example, monoamine oxidase inhibitor antidepressants cause
a decreased metabolism of dietary amines, such as tyramine, and this has
caused serious and sometimes fatal interactions when cheese is eaten (see
GRIFFIN et al. 1988).
Cimetidine is known to decrease the clearance of other drugs that are
largely eliminated by hepatic metabolism. It does this mainly by binding to the
hepatic microsomal cytochrome P450 mixed-function oxidase system, thus
inhibiting the metabolism of other drugs. However, with cimetidine other
mechanisms may be involved since it also reduces liver blood flow and this
reduced perfusion may potentiate the action of drugs (e.g., propranolol)
whose systemic clearance is highly dependent upon liver blood flow.
WOLFF et al. (1993) have emphasised that more knowledge about struc-
tural features of compounds inhibitory for cytochrome P450 enzymes is re-
quired to evaluate whether one drug may be inhibitory for the metabolism of
one or more drugs. They have suggested and described two approaches to
elucidate the specificity of cytochrome P450 enzymes on a molecular level:
firstly, to determine the three-dimensional structure of the active site by X-ray
crystallography and use of this information for the development of substrate
and inhibitor models and, secondly, the design of pharmacophor models of
substrates and inhibitors on the basis of quantitative structure-activity rela-
tionships or by using molecular modelling methods.

F. Conclusions
It may be apparent from these brief accounts of interactions involving P450
enzyme induction or inhibition that many of the involved drugs are those on
which patients are carefully stabilised often for long periods of time (e.g.,
anticoagulants, antidepressants, hypoglycaemics, hypotensives, oral contra-
ceptives). Past experience has shown that it is these drug-stabilised patients
who are at special risk from any interaction that will influence the potency or
availability of their medication. This is especially so for the elderly patients,
whose medication needs are generally greater than younger persons and who
have a substantially greater risk than the younger patient of experiencing
adverse reactions to medication.
Drug Interactions and Drug-Metabolising Enzymes 165

In discussing the role of cytochrome P450 enzymes in the mechanisms of

these interactions some caveats must be introduced about the information
that has been used to establish their veracity. Much of the current knowledge
of the cytochrome P450 enzyme systems has been gained from studies in
animals. However, animal species differ from each other and from humans
in respect to their pattern of cytochrome P450 enzymes involved in detoxifica-
tion or bioactivation of xenobiotics, and cytotoxic and genotoxic chemicals.
It is important to know the specificity of these enzymes and their distribution
among human and animal species if animals studies are to be useful in
understanding actual interactions or predicting likely interactions between
therapeutic drugs.
It is also tempting to focus all attention on the role of the cytochrome
P450 enzymes; however, this enzyme system is but one of the many that
may exist in the cell for the metabolism of xenobiotics. Cytochrome P450 has
been the easiest to study because it is a coloured protein that can readily be
evaluated spectrophotometric ally, but it is not all of drug metabolism
and other enzyme systems do exist which should not be neglected if a full
understanding is to be gained of how drugs function as enzyme inducers
(ESTABROOK 1979).
Although the majority of oxidation reactions of xenobiotics are catalysed
by the cytochrome P450 monooxygenase system (NEBERT and GONZALEZ
1985), certain oxidative reactions of xenobiotics are also catalysed by enzymes
other than the cytochrome P450 monooxygenase system, for example the
microsomal FAD-containing monooxygenase (ZEIGLER 1984, 1985; DAMANI
1988). This is responsible for the N-oxidation of tertiary amines such as
dimethylaniline and trimethylamine. The enzyme requires NADPH and
oxygen; the substrate specificity also includes secondary amines and
sulphides, thioethers and thiocarbamates, and organophosphates (TIMBRELL
1993). Alcohols may be oxidised by alcohol dehydrogenase; xanthine
oxidase catalyses the oxidation of nitrogen heterocyclics such as the purine
hypoxanthine. Some amines such as tyramine are substrates for monoamine
oxidases; diamines such as putrescine are metabolised by the soluble enzyme
diamine oxidase. The peroxidases are also involved in the oxidation of
xenobiotics, the most important being prostaglandin synthase, which is known
to catalyse the oxidation of p-phenetidine, a metabolite of phenacetin, a
process that may be involved in the nephrotoxicity of the drug (TIMBRELL
1993).
The role of enzyme systems in the metabolism of therapeutic agents is
indeed a fascinating study; the knowledge base is ever growing and the genetic
basis of enzyme source and action is gradually being revealed. Patient therapy
can only benefit from these disclosures, and safety in therapy is one of the
prime contributions that such knowledge may provide. Much of the empirical
approach to therapy is being replaced by judgements based upon proven fact
and one might well look forward to the future when synergistic drug-drug
interactions are not hazardous but only beneficial.
166 P.F. D'ARCY

References
Acocella G, Bonollo L, Garimoldi M, Mainardi M, Tenconi LT, Nicolis FB (1972)
Kinetics of rifampicin and isoniazid administered alone and in combination to
normal subjects and patients with liver disease. Gut 13:47-53
Adebayo GI (1988) Interaction between phenytoin and theophylline in healthy volun-
teers. Clin Exp Pharmacol Physiol 15:883-887
Ahmad SR (1992) USA antihistamine alert. Lancet 340:542
Alvares AP, Schilling G, Levin W, Kuntzman R (1967) Studies on the induction of CO-
binding pigments in liver microsomes by phenobarbital and 3-methylcholan-
threne. Biochem Biophys Res Commun 29:521-526
Angle M, Huff PS, Lea JW (1991) Interactions between oral contraceptives and thera-
peutic drugs. Outlook 9:1-6
Arin,< E (1993) Extrahepatic microsomal forms: lung microsomal cytochrome P450
isozymes. In: Schenkman JB, Greim H (eds) Cytochrome P450. Springer, Berlin
Heidelberg New York, pp 373-386
Armstrong G, Beg MF, Scahill S (1991) Warfarin potentiated by proguanil. Br Med J
3034:789
Axelrod J (1955) The enzymatic demethylation of ephedrine. J Pharmacol Exp Ther
114:430-438
Ayesh R, Idle JR, Ritchie JC, Crothers MJ, Hetzel MR (1984) Metabolic oxidation
phenotypes as markers for susceptibility to lung cancer. Nature 311:169-170
Barbeau A, Cloutier T, Roy M, Plasse L, Paris S, Poirier J (1985) Ecogenetics of
Parkinson's disease: 4-hydroxylation of debrisoquine. Lancet 11:1213-1216
Bertilsson L, Dengler HJ, Eichelbaum M, Schultz HU (1980) Pharmacogenetic
covariation of defective N-oxidation of sparteine and 4-hydroxylation of
debrisoquine. Eur J Clin PharmacoI17:153-155
Birnbaum LS (1993) Changes in cytochrome P450 in senescence. In: Schenkman JB,
Greim H (eds) Cytochrome P450. Springer, Berlin Heidelberg New York, pp 477-
492
Black SD (1993) Cytochrome P450 structure and function. In: Schenkam JB, Greim H
(eds) Cytochrome P450. Springer, Berlin Heidelberg New York, pp 156-168
Booth CD (1993) A drug interaction between danazol and warfarin. Pharm J 250:439-
440
Brodie BB, Axelrod J, Cooper JR, Gaudette LE, La Duc BN, Mitowa C, Udenfriend
S (1955) Detoxification of drugs and other foreign compounds by liver mi-
crosomes. Science 121:603-604
Brooks BA, McBride OW, Dolphin CT, Farrall M, Scambler PJ, Gonzalez FJ, Idle JR
(1988) The gene CYP3 encoding P45PCN1 (nifedipine oxidase) is tightly linked to
the gene COLIA2 encoding collagen type I alpha on 7q21-p22.1. Am J Hum Genet
43:280-284
Brosen K, Gram LF (1989) Clinical significance of the sparteine/debrisoquine oxida-
tion polymorphism. Eur J Clin PharmacoI36:537-547
Cadle RM, Zenon GJ III, Rodriguez-Barradas MC, Hamill RJ (1994) Fluconazole
induced symptomatic phenytoin toxicity. Ann Pharmacother 28:191-195
Cashman JR, Hanks D, Weiner RI (1987) Epoxy derivatives of arachidonic acid are
potent stimulators of prolactin secretion. Neuroendocrinology 46:245-251
Castel JM, Cappiello E, Leopaldi D, Latini R (1990) Rifampicin lowers plasma
concentrations of propafenone and its antiarrhythmic effect. Br J Clin Pharmacol
30:155
Committee on Safety of Medicines (1989) Serious interaction between tamoxifen and
warfarin. Committee on Safety of Medicines, London (Current problems, no 26)
Conney AH (1967) Pharmacological implications of microsomal enzyme induction.
Pharmacol Rev 19:317-366
Conney AH, Miller EC, Miller JA (1956) The metabolism of methylated aminoazo
dyes. V. Evidence for induction of enzyme synthesis in the rat by 3-methylcholan-
threne. Cancer Res 16:450-459
Drug Interactions and Drug-Metabolising Enzymes 167

Conney AH, Miller EC, Miller JA (1957) Substrate-induced synthesis and other
properties of benzpyrenehydroxylase in rat liver. J BioI Chern 228:753-766
Consumers' Association (1993) Fluoroquinolones reviewed. Drug Ther Bull
31(18):69-72
Crespi CL, Penman BW, Gelboin HV, Gonzalez FJ (1991) A tobacco smoke-derived
nitrosamine 4-(methylnitrosamino)-I-(3-pyridyl)-I-butanone is activated by
multiple human cytochrome P450s including the polymorphic human cytochrome
P4502D6. Carcinogenesis 12:1197-1201
Daly AK, Idle JR (1993) Genetics: animal and human cytochrome P450 polymor-
phisms. In: Schenkman JB, Grein H (eds) Cytochrome P450. Springer, Berlin
Heidelberg New York, pp 433-446
Damani LA (1988) The flavin-containing monooxygenase as an amine oxidase. In:
Gorrod JW, Oelschlager H, Caldwell J (eds) Metabolism of xenobiotics. Taylor
and Francis, London, pp 59-70
D'Arcy PF (1984a) Tobacco smoking and drugs: a clinically important interaction?
Drug Intell Clin Pharm 18:302-307
D'Arcy PF (1984b) Vaccine-drug interactions. Drug Intell Clin Pharm 18:697-700
D'Arcy PF (1986) Drug interactions with oral contraceptives. Drug Intell Clin Pharm
20:353-362
D'Arcy PF, Merkus FWHM (1981) Alcohol and drug interactions. Pharm Int 2:273-
280
Davis GC, Williams AC, Markey SP, Ebert MH, Caine ED, Reichert CM, Kopin IJ
(1979) Chronic parkinsonism secondary to intravenous injection of meperidine
analogues. Psychiatry Res 1:249-254
Dayer P, Balant L, Courvoisier F, Kupfer A, Kubli A, Georgia A, Fabre J (1982) The
genetic control ofbufuralol metabolism in man. Eur J Drug Metab Pharmacokinet
7:73-77
Dean WL, Gray RD (1982) Relationship between state of aggregation and catalytic
activity for cytochrome P-450 LM2 and NADPH-cytochrome P-450 reductase.
J BioI Chern 257:14679-14685
DeWaziers I, Cugnenc P, Yang CS, Leroux JP, Beaune P (1990) Cytochrome P-450
isozymes, epoxide hydrolase and glutathione transferase in rat and human hepatic
and extrahepatic tissues. J Pharmacol Exp Ther 253:387-394
Dillard ML, Fink RM, Parkerson R (1992) Ciprofloxacin-phenytoin interaction. Ann
Pharmacother 26:263
Eichelbaum M, Spannbrucker N, Steincke B, Dengler HJ (1979) Defective N-oxidation
of sparteine in man: a new pharmacogenetic defect. Eur J Clin PharmacoI17:153-
155
Elofsson S-A, Gamberale F, Hindmarsh T, Iregren A, Isaksson A, Johnsson I, Knave
B, Lydahl E, Mindus P, Persson HE, Philipson B, Steby M, Struwe G, SOderman
E, Wennberg A, Widen L (1980) Exposure to organic solvents. A cross-sectional
epidemiological investigation on occupationally-exposed car and industry spray
painters with special reference to the nervous system. Scand J Work Environ
Health 6:239-273
Estabrook RW (1979) Concluding remarks. In: Estabrook RW, Lindenlaub E (eds)
The induction of drug metabolism. Schattauer, Stuttgart, pp 643-645 (Symposia
Medica Hoechst, vol 14)
Evans DAP, Mahgoub A, Sloan TP, Idle JR, Smith RL (1980) A family and population
study of the genetic polymorphism of debrisoquine oxidation in a British white
population. J Med Genet 17:102-105
Gannon RH, Anderson ML (1992) Fluconazole-nortriptyline drug interaction. Ann
Pharmacother 26:1456-1457
Gelehrter TD (1979) Enzyme induction in mammals - an overview. In: Estabrook RW,
Lindenlaub E (eds) The induction of drug metabolism. Schattauer, Stuttgart, pp 7-
24
Gonzalez FJ (1993) Cytochrome P450 in humans. In: Schenkman JB, Greim H (eds)
Cytochrome P450. Springer, Berlin Heidelberg New York, pp 239-257
168 P.F. D'ARCY

Gorrie M, Beaman M, Nicholls A, Blackwell P (1994) Allopurinol interaction with

cyclosporin. Br Med J 308:113
Griffin JP, D' Arcy PF, Speirs CJ (1988) A manual of adverse drug interactions, 4th edn.
Wright, London
Guengerich FP, Kim DH (1990) In vitro inhibition of dihydropyridine oxidation and
aflatoxin Bl activation in human liver microsomes by naringin and other
flavonoids. Carcinogenesis 11 :2275-2279
Guengerich FP, Marin MV, Beaune PH, Kremers P, Wolff T, Waxman DJ (1986)
Characterization of rat and human liver microsomal cytochrome P-450 forms
involved in nifedipine oxidation, a prototype for genetic polymorphism in oxida-
tive drug metabolism. J BioI Chern 261:5051-5060
Gupta A, Sexton RC, Rudney H (1986) Modulation of regulatory oxysterol formation
and low density lipoprotein suppression of 3-hydroxy-3-methyl glutaryl coenzyme
A (HMG-Coa) reductase activity by ketoconazole. A role for cytochrome P450 in
the regulation of HMG-Coa reductase in rat intestinal epithelia cells. J BioI Chern
261:8348-8356
Horowitz HW, Jorde UP, Wormser GP (1992) Drug interactions in use of dapsone for
Pneumocystis carinii prophylaxis. Lancet 339:747
Hoyumpa AM Jr, Schenker S (1982) Major drug interactions: effect of liver disease,
alcohol and nutrition. Ann Rev Med 33:113-149
Jassal SV (1991) Warfarin potentiation by glibenclamide. Br Med J 303:789
Kolmodin B, Azamoff DL, Sjoqvist F (1969) Effect of environmental factors on drug
metabolism: decreased plasma half-life of aminopyrine in workers exposed to
chlorinated hydrocarbon insecticides. Clin Pharmacol Ther 10:638-642
Langhoff E, Madsen S (1983) Rapid metabolism of cyclosporin and prednisone in
kidney transplant patients receiving tuberculostatic treatment. Lancet 11:1031
Le Goascogne C, Robel P, Gouezow M, Snanes N, Baulieu E-E, Waterman M (1987)
Neurosteroids: cytochrome P450 in the rat brain. Science 237:1212-1214
Lindell TJ, Ellinger R, Warren JT, Sundheimer D, O'Malley AF (1977) The effect
of acute and chronic phenobarbital treatment on the activity of rat liver
deoxyribonuclease acid-dependent ribonucleic acid polymerases. Mol Pharmacol
13:426-434
Lodwick R, McConkey B, Brown AM (1987) Life threatening interaction between
tamoxifen and warfarin. Br Med J 295:1141
Lu A YH, Coon MJ (1968) Role of hemoprotein P450 in fatty acid w-hydroxylation in
a soluble enzyme from liver microsomes. J BioI Chern 243:1331-1332
Mahgoub A, Idle JR, Dring LG, Lancaster R, Smith RL (1977) Polymorphic hydroxy-
lation of debrisoquine in man. Lancet 11:584-586
Maizlish NA, Langolf GD, Whitehead LW, Fine LJ, Alberts JW, Goldberg J, Smith P
(1985) Behavioral evaluation of workers exposed to mixtures of organic solvents.
Br J Ind Med 43:257-262
Mallov JS (1976) MBK neuropathy among spray painters. JAMA 235:1455-1457
Maltony C, Ciliberti A, Caretti D (1982) Experimental contributions in identifying
brain potential carcinogens in the petrochemical industry. Ann NY Acad Sci
381:216-249
Matsubara S, Yamamoto S, Sogawa K, Yokotani N, Fujii-Kuriyama Y, Haniu M,
Shively JE, Gotoh 0, Kusunose E, Kusunose M (1987) cDNA cloning and
inducible expression during pregnancy of the mRNA for rabbit pulmonary
prostaglandin w-hydroxylase (cytochrome P-450P2). J BioI Chern 262:13366-
13371
McFadden JP, Pontin JE, Powles A V, Fry L, Idle JR (1989) Cyclosporin decreased
nifedipine metabolism. Br Med J 299:1224
McManus ME, Boobis AR, Pacifici GM, Frempang RY, Brodie MJ, Kahn GC, Whyte
C, Davies DS (1980) Xenobiotic metabolism in the human lung. Life Sci 26:481-
487
Meeks ML, Mahaffey KW, Katz MD (1992) Danazol increases the anticoagulant
effects of warfarin. Ann Pharmacother 26:641-642
Drug Interactions and Drug-Metabolising Enzymes 169

Monig H, Baese C, Heidemann HT, Ohnhaus EE, Schulte HM (1990) Effect of oral
contraceptive steroids on the pharmacokinetics of phenprocoumon. Br J Clin
PharmacoI30:115-116
Mi.iller GC, Miller JA (1953) The metabolism of methylated aminoazo dyes. II.
Oxidative demethylation by rat liver homogenates. J BioI Chern 202:579-
587
Murphy RC, Falk JR, Lumni S, Yadagari P, Zirrolli JA, Balazy M, Masferrer JL,
Abraham NG, Schwartzman ML (1988) 12(R)-Hydroxyeicosatrienoic acid: a
vasodilator cytochrome P450-dependent arachidonate metabolite from bovine
corneal epithelium. J BioI Chern 263:17197-17202
Nebert DW, Gonzalez EJ (1985) Cytochrome P-450 gene expression and regulation.
Trends Pharmacol Sci 6:160-164
Noticeboard (1992) Warning about astemizole. Lancet 340:1155
Oates S, Shah RR, Idle JR, Smith RL (1982) Genetic polymorphism of phenformin 4-
hydroxylation. Clin Pharmacol Ther 32:81-89
Pantuck EJ, Hsiao KC, Kuntzman R, Conney AH (1975) Intestinal metabolism of
phenacetin in the rat: effect of charcoal-broiled beef and rat chow. Science
187:744-745
Parke DV (1975) Induction of the drug-metabolising enzymes. In: Parke DV (ed)
Enzyme-induction. Plenum, London, p 207
Parke DV (1979) The responsiveness of cells to various drug inducers. In: Estabrook
RW, Lindenlaub E (eds) The induction of drug metabolism. Schattauer, Stuttgart,
pp 105-111
Philpot RM, Wolf CR (1981) The properties and distribution of the enzymes of pulmo-
nary cytochrome P450 dependent monooxygenase systems. In: Hodgson E, Bend
JR, Philpot RM (eds) Reviews in biochemical toxicology, vol 3. Elsevier,
Amsterdam, pp 51-76
Philpot RM, Arin,< E, Fouts JR (1975) Reconstitution of the rabbit pulmonary microso-
mal mixed-function oxidase system from solubilized components. Drug Metab
Dispos 3:118-126
Posner HS, Mitoma C, Udenfriend S (1961) Enzymatic hydroxylation of aromatic
compounds. II. Further studies of the properties of the microsomal hydroxylating
system. Arch Biochem Biophys 94:269-279
Remmer H, Merker HJ (1963) Drug induced changes in the liver endoplasmic
reticulum: association with drug metabolizing enzymes. Science 14:1657-
1658
Rennert H, Fischer RT, Alkvarez JG, Trzaskos JM, Strauss JM (1990) Generation of
regulatory oxysterols: 26-hydroxylation of cholesterol by ovarian mitochondria.
Endocrinology 127:738-746
Sacerdoti D, Abraham NG, McGiff JC, Schwartzman ML (1988) Renal cytochrome
P450-dependent metabolism of arachidonic acid in spontaneously hypertensive
rats. Biochem PharmacoI37:521-527
Schenkman JB (1993) Historical background and description of the cytochrome P450
monooxygenase system. In: Schenkman JB, Greim H (eds) Cytochrome P450.
Springer, Berlin Heidelberg New York, pp 3-13
Schmucker DL, Woodhouse KW, Wang RK, Wynne H, James OF, McManus M,
Kremers P (1990) Effects of age and gender on in vitro properties of human liver
microsomal monooxygenases. Clin Pharmacol Ther 48:365-374
Seaton TL, Ceklum CL, Back DJ (1991) Possible potentiation of warfarin by
fluconazole. Ann Pharmacother 24:1177-1178
Serpa MD, Cossolias J, McCreevy MJ (1992) Moricizine-warfarin: a possible drug
interaction. Ann Pharmacother 26:127
Shaeffer MS, Collier D, Sorrell MF (1994) Interaction between FK 506 and erythromy-
cin. Ann Pharmacother 28:280-281
Shimada T, Guengerich FP (1989) Evidence for cytochrome P-450NF, the nifedipine
oxidase, being the principal enzyme involved in the bioactivation of aflatoxins in
human liver. Proc Natl Acad Sci USA 86:462-465
170 P.F. D'ARCY

Sladek NE, Mannering GJ (1966) Evidence for a new P-450 hemoprotein in hepatic
microsomes from methylcholanthrene treated rats. Biochem Biophys Res
Commun 24:668-674
Snyder GD, Capdevila J, Chacos N, Manna S, Flack JR (1983) Action of luteinizing
hormone-releasing hormone: involvement of novel arachidonic acid metabolites.
Proc Nat! Acad Sci USA 80:3504-3507
Soto J, Sacrisan JA, Arellano F, Mazas J (1991) Possible theophylline-amiodarone
interaction. Ann Pharmacother 24:1115
Srivastava SP, Seth PK (1983) 7-Ethoxycoumarin O-deethylase activity in rat brain
microsomes. Biochem Pharmacol 32:3657-3660
Staum JM (1990) Enzyme induction: rifampicin-disopyramide interaction. Ann
Pharmacother 24:701-703
Strobel HW, Stralka DJ, Hammond DK, White T (1993) Extrahepatic microsomal
forms: gastrointestinal cytochromes P450, assessment and evaluation. In:
Schenkman JB, Greim H (eds) Cytochrome P450. Springer, Berlin Heidelberg
New York, pp 363-371
Tenni P, Lalich DL, Byrne MJ (1989) Life threatening interaction between tamoxifen
and warfarin. Br Med J 298:93
Timbrell JA (1993) Biotransformation of xenobiotics. In: Ballantyne B, Marrs T,
Turner P (eds) General and applied toxicology, voll. Macmillan, Basingstoke, pp
89-119
Tisdel KA, Israel D SA, Kolb KW (1994) Warfarin-felbamate interaction: first report.
Ann Pharmacother 28:805
Tjia JF, Back DJ, Breckenridge AM (1989) Calcium channel antagonists and
cyclosporin metabolism: in vitro studies with human liver microsomes. Br J Clin
PharmacoI28:362-365
Volk B, Amelized Z, Anagnostopoulos J, Knoth R, Oesch F (1988) First evidence of
cytochrome P450 induction in the mouse brain by phenytoin. Neurosci Lett
84:219-224
Warner A, Gustafsson J-A (1993) Extrahepatic microsomal forms: brain cytochrome
P450. In: Schenkman JB, Greim H (eds) Cytochrome P450. Springer, Berlin
Heidelberg New York, pp 387-397 0

Warner M, Toilet P, Carlstrom K, Gustafsson J-A (1989) Endocrine regulation

of cytochrome P450 in the rat brain and pituitary gland. J Endocrinol 122:341-
349
Watkins PB, Wrighton PB, Schuetz EG, Molowa DT, Guzelian PS (1987) Identifica-
tion of glucocorticoid-inducible cytochromes PA50 in the intestinal mucosa of rats
and man. J Clin Invest 80:1029-1036
Wattenberg LW, Leon JL (1962) Histochemical demonstration of reduced pyridine
nucleotide dependent polycyclic hydrocarbon metabolizing systems. J Histochem
Cytochem 10:412-420
Wattenberg LW, Leong JL, Strand PJ (1962) Benzpyrene hydroxylase activity in the
gastrointestinal tract. Cancer Res 22:1120-1125
Williams DE, Hale SE, Okita RT, Masters BSS (1984) A progtaglandin co-hydroxylase
cytochrome PA50 (P-450PGw) purified from lungs of pregnant rabbits. J Bioi
Chern 259:14600-14608
Williams RT (1959) Detoxification mechanisms. Cunningham, London
Wolff T, Strobl G, Greim H (1993) Structural models for substrates and inhibitors of
cytochrome P450 enzymes. In: Schenkman JB, Greim H (eds) Cytochrome P450.
Springer, Berlin Heidelberg New York, pp 195-207
Woodcock BG, Kirsten R, Nelson K, Rietbrock S, Hopf R, Kaltenbach M (1991)
A reduction in verapamil concentrations with phenytoin. N Engl J Med 325:
1179
Zeigler DM (1984) Metabolic oxygenation of organic nitrogen and sulphur com-
pounds. In: Mitchell JR, Horning MG (eds) Drug metabolism and drug toxicity.
Raven, New York, pp 33-53
Drug Interactions and Drug-Metabolising Enzymes 171

Zeigler DM (1985) Molecular basis for N-oxygenation of sec- and tert-amines. In:
Gorrod JW, Damani LA (eds) Biological oxidation of nitrogen in organic mol-
ecules. Ellis Horwood, Chichester, pp 45-52
Zornberg GL, Bodkin JA, Cohen BA (1991) Severe adverse interactions between
pethidine and selegiline. Lancet 337:246
CHAPTER 6
Drug Interactions
Involving Renal Excretory Mechanisms
A. SOMOGYI

A. Introduction
The kidney represents the final elimination organ for virtually all foreign
substances irrespective of whether they are cleared unchanged by the kidney
or whether as metabolites formed in the body (predominantly the liver). It had
long been regarded that filtration at the glomerulus was the sole excretory
mechanism in the kidney for foreign substances. However, in 1874,
Heidenhain demonstrated the phenomenon of secretion by canine tubular
cells using the dye indigo carmine, and Marshall and colleagues in the 1920s
conclusively showed secretion of organic anions by the proximal tubules. Thus
the concept of filtration and secretion by the kidney came into existence. In
1929 M~ller and co-workers coined the term clearance, in which they math-
ematically described the excretion of substances by the kidney. It is
somewhat surprising that the concepts developed by these early pioneers of
renal physiology are still highly applicable in pharmacokinetics today. The
active tubular secretion of organic cations was shown for the first time by
Rennick in the dog using tetraethylammonium and by Sperber using a variety
of organic cations in the chicken. Further details of the history of renal physi-
ology/pharmacology can be found in several reviews (PETERS 1960; WEINER
and MUDGE 1964; RENNICK 1972).

B. Mechanisms of Renal Excretory Clearance

I. Anatomy and Physiology of the Kidney

The major function of the kidney is to remove waste products from the blood
and to regulate fluid and electrolyte content in the body. The functional unit of
the kidney is the nephron, which filters, secretes and reabsorbs molecules.
Blood supply to the kidneys is via the left and right renal arteries at the rate of
approximately 1200ml/min or 20%-25% of cardiac output.

II. Glomerular Filtration

Filtration involves the forcing of fluid containing dissolved substances through
the endothelial-capsular membrane. This membrane, which acts as a filter,
174 A. SOMOGYI

allows some materials to pass through whilst restricting the passage of others.
The resulting filtrate is free of protein and substances with a molecular weight
of greater than 40000 and about 20A in diameter. The rate of glomerular
filtration is about 1801/day or 120ml/min.
The pharmacokinetic consequence of glomerular filtration is that the
clearance rate of drugs at the glomerulus [CL(GFR)] can be approximated to
the unbound fraction (fu) in plasma of the drug multiplied by the glomerular
filtration rate (GFR).
CL(GFR) = fu x GFR
Hence for drugs which are highly bound (fu < 0.1) in plasma, clearance at the
glomerulus is substantially less than the glomerular filtration rate, and the
maximum clearance of any drug at the glomerulus will be equal to the glom-
erular filtration rate.
The mechanisms by which drugs can interact at the glomerulus are: (a)
displacement from protein-binding sites. The pharmacokinetic consequence
will be that fu will increase and clearance of the drug by the glomerulus will
also increase in direct proportion to the increase in fu. There are very few
examples in the literature of drug-drug interactions occurring exclusively at
the glomerulus. (b) Where the glomerulus is physically damaged through the
nephrotoxic action ofaxenobiotic. In this case, the maximum clearance of a
drug will be equal to the GFR, irrespective of the value of the drug's fu.
However, this type of interaction is poorly documented and would rarely
occur on its own.

III. Tubular Secretion

The blood flow to the kidney then passes to the tubular network, where active
transport systems (requiring energy) are capable of removing drugs and en-
dogenous substances from blood and secreting them into the lumen of the
kidney. This occurs at the site of the proximal tubule, where there are a
number of transporters for organic anions and organic cations. There are
specific transporters for organic cations and organic anions at both the luminal
(brush border) and contraluminal membranes of the epithelial cells lining the
proximal tubule.

1. Physiological Considerations
The blood flow to the proximal tubule via the renal arteries is approximately
85% of total renal blood flow. Thus the maximum clearance of any drug
by tubular secretion at this site is equivalent to the total renal blood flow of
lo21/min.

2. Cellular Mechanisms
Our increasing understanding of the cellular mechanisms of transport of drugs
across the epithelial membranes of the proximal tubule has been accomplished
Drug Interactions Involving Renal Excretory Mechanisms 175

by the use of a number of techniques such as isolated perfused and non-

perfused tubules, isolated perfused kidneys, plasma membrane vesicles and
primary or continuous kidney cell lines. The results of transport studies using
these techniques have allowed prediction of drug interactions in humans
(SOMOGYI et al. 1989).

a) Transport Processes for Organic Anions

This has been reviewed by several groups (M0LLER and SHEIKH 1983; ULLRICH
1994). Although not fully evaluated in humans, the renal transporter of or-
ganic anions (p-aminohippuric acid being the most commonly used probe
substrate) is located at the contraluminal cell membrane (Fig. 1). The cellular
uptake of organic anions has been termed a tertiary active transport process,
since the primary transporter is the ATP sodium/potassium transport step,
which activates a sodium-dicarboxylate transporter, creating an a-
ketoglutarate gradient which is exchanged for the organic anion. At the lumi-
nal membrane, the transport mechanism(s) has not been identified. The
specificity of the contraluminal transporter has been summarized by ULLRICH
(1994); the substrate must have a hydrophobic domain of greater than 4A; the
strength of the ionic charge on the substrate is a determinant of the interaction
with the transporter; the transporter also accepts hydrophobic substrates and,
finally, electron-attracting side groups (such as CI, Br) augment the interaction
of the substrate with the transporter.

b) Transport Processes for Organic Cations

This has been reviewed on a number of occasions (PETERS 1960; RENNICK
1981). Secretion of organic cations across the mammalian kidney occurs along

Proximal Tubular Cell

Lumen
2Na+ Sulfate
Sulfate Oxalate
3Na+ 3Na+
Dicarboxylates Dicarboxylates
PAH Hydrophobic
Monocarboxylate (PAH)
X
a - Ketoglutarate
2Na+
Lactate
N i-Methyl-nicotinamide
H+ (NMeN+)
Cations X

Luminal Transporters Contraluminal Transporters

Fig. 1. Location of transporters for organic anions and cations in the renal proximal
tubule (from ULLRICH 1994). Note: PAH is the organic anion p-aminohippuric acid
176 A. SOMOGYI

the entire length of the proximal tubule. At the contraluminal membrane

there is evidence for one common transport system. At the luminal membrane,
the transport of organic cations occurs by an electroneutral H+/organic cation
system. The driving force at this luminal interface is the sodium/hydrogen
transporter, which transports sodium ions into the cell and hydrogen ions out
(Fig. 1). More than one transport system exists for organic cations at the
luminal membrane, since choline has a high affinity for a transporter which is
not H+ driven. The key substrate requirements are similar to those of the
contraluminal organic anion transporter (hydrophobicity, ionic charge
strength and hydrogen bond formation). Substrates invariably contain a nitro-
gen atom (ULLRICH 1994). The cloning and sequencing of such organic cation
transporters is only in its infancy but will provide further valuable information
on substrate specificity.

3. Pharmacokinetic Evidence for Tnbular Secretion

To determine whether a drug in vivo undergoes net active tubular secretion,
knowledge of the value of its renal excretory clearance is required. This is
usually obtained from pharmacokinetic studies in which plasma concentra-
tions and the amount of drug recovered in urine as the unchanged form of
the drug are quantified. Renal clearance (CLR ) is calculated as the amount
excreted in urine divided by the area under the plasma concentration-
time curve. When this value is greater than clearance by the glomerulus
[CL(GFR); see equation above], active secretion is inferred. That is, if C~ >
CL(GFR) then the drug is secreted by the proximal tubules. For example,
cimetidine has a total body clearance of 600mllmin and 70% of the dose is
recovered unchanged in urine. Hence its renal clearance is 0.7 x 600 = 420mll
min. This is greater than clearance at the glomerulus [0.8(fu) x 120 = 96mll
min], allowing one to conclude that cimetidine also undergoes net renal tubu-
lar secretion.
This knowledge allows one to examine the sites in the kidney for the renal
elimination of drugs. In terms of renal tubular secretion, when renal clearance
indicates tubular secretion then the chemical nature of the drug needs to be
ascertained. If the drug is an organic anion then one can assume that the drug
is secreted by the organic anion transporter and if an organic cation then it is
secreted by the organic cation transporter. A key feature of a drug undergoing
tubular secretion is that it will compete with another drug for secretion. In all
studies to date there is little evidence that competition for transport is non-
competitive but rather competitive. Thus the drug with the higher affinity for
the transporter will be a more potent inhibitor of the transport of a similar
chemical substance for the same transporter. Hence organic anions will com-
pete with one another for transport (e.g., probenecid and penicillin) and
organic cations will compete with one another (e.g., cimetidine and
procainamide). This is the most common type of drug-drug interaction involv-
ing renal excretory mechanisms and will form the basis of this review.
Drug Interactions Involving Renal Excretory Mechanisms 177

IV. Tubular Reabsorption

As drug passes down the tubular lumen into the distal tubule and collecting
duct, passive diffusion (driven by the concentration gradient) of drug across
the cellular membrane back into the blood can occur. This process of passive
reabsorption is applicable to the unionized lipid-soluble form of the drug.
Thus the pKa of the drug and the pH of the tubular fluid (urine) become
important, as only the unionized form is sufficiently lipid soluble to diffuse
through the epithelial membranes of the tubule cells. Hence any drug which
alters urine pH or urine flow rate (which alters the concentration gradient) has
the potential for altering the tubular reabsorption of other drugs. The signifi-
cance of any interaction at this site is often of minor importance since the basic
requirement is the degree of lipid solubility of the drug and in general lipid-
soluble drugs are extensively metabolized by the liver rather than being exten-
sively cleared by the kidney. The few exceptions include the ephedrine series
of drugs, salicylate and chlorpropamide.

C. Drug Interactions Involving Tubular Secretion

Of the mechanisms for renal drug excretion, drug interactions predominantly
involve tubular secretion. Many in vivo studies in humans have been con-
ducted characterizing interactions involving drugs which are either organic
anions or organic cations.

I. Organic Anions
The most commonly investigated interacting drug has been probenecid with
many published studies showing that it reduces the renal clearance of drugs
which are organic anions.

1. Probenecid
It was the discovery of penicillin by Lord Florey which ultimately led to the
discovery of probenecid. Penicillin although effective had a short duration of
action due to its very high renal clearance and short half-life. A number of
strategies were investigated to overcome this therapeutic problem. One strat-
egy was to try and block the renal excretion of penicillin. The development of
probenecid (BEYER et al. 1951) heralded the clinical practice of elevating blood
concentrations of organic anion drugs by blocking their renal tubular secre-
tion. Since then, probenecid has been used as a prototypic in vitro and in vivo
probe to assess whether an organic anion drug is secreted by the proximal
tubules. This has been documented for a number of drugs but, in particular,
penicillin and cephalosporin antibiotics. In this review, the symbol ± always
refers to the standard deviation.
178 A. SOMOGYI

a) Allopurinol
EUON et al. (1968) showed in three patients with gout that the administration
of 1 g probenecid increased the renal clearance of oxypurinol (the active
metabolite of allopurinol) from about 12ml/min to 36ml/min, without any
change in inulin clearance (a measure of GFR). The mechanism for this
increase in renal clearance involves inhibition by probenecid of the active
luminal re-uptake of oxypurinol, via the uric acid transport system.

b) Enprofylline and Dyphylline

In six normal healthy subjects, 1 g probenecid reduced the renal clearance of
enprofylline, administered as a 1-mg/kg intravenous infusion, from 17 ± S1/h to
8 ± 31/h (P < 0.01) (BORGA et al. 1986), with a reduction in total body clearance
and increase in half-life. Similarly, in 12 healthy subjects, the total body clear-
ance of dyphylline (a dihydroxypropyl derivative of theophylline) was reduced
(P < 0.001) by almost SO% from 178 ± 8 to 101 ± 7mllh per kilogram by 19
probenecid (MAY and JARBOE 1981). Enprofylline and dyphylline are organic
acid methylated xanthines and have renal clearance values greater than the
glomerular filtration rate, indicating net tubular secretion.

c) Frusemide
Frusemide is an organic acid, highly protein bound (unbound fraction is less
than 0.1), loop diuretic which produces its pharmacological effect(s) at the
luminal rather than the peritubular site of the renal tubule. Access to the
luminal site is mainly by active secretion by the organic anion transport sys-
tem. It was hypothesized that administration of probenecid could blunt the
diuretic and natriuretic effect of frusemide by blocking access to its site of
action. Four healthy normal subjects received a single intravenous injection of
1 mg/kg frusemide with and without pretreatment of O.S g probenecid 8 and
2 h before the frusemide dose. Probenecid reduced the renal clearance of
frusemide from 134 ± 23 to 63 ± 6mllmin (P < O.OS). However, there was no
significant difference in the 6-h urine volume or sodium excretion (HONARI et
al. 1977). When frusemide was administered as an intravenous infusion in four
subjects, the reduction in frusemide renal clearance was accompanied by a
rise in plasma frusemide concentrations, but urine flow rate (29 ± 4 ml/min
frusemide alone and 15 ± 4mllmin with probenecid) and excreted fraction of
filtered sodium (19% ± 7% alone and 14% ± 9% with probenecid) declined
significantly (P < 0.05) (HONARI et al. 1977). In six normal, healthy subjects, the
effect of probenecid (0.5 g every 6 h for 3 days) on the disposition and effect
of frusemide 40mg intravenously was evaluated (HOMEIDA et al. 1977).
Probenecid significantly (P < 0.01) reduced the renal clearance of frusemide
from 90 ± 24 to 20 ± 12ml/min. In the first 30min after frusemide administra-
tion, there was a significant (P < 0.01) reduction in sodium excretion rate from
3.1 ± 0.6 to 2.2 ± 0.7mmollmin. However, there was no difference in the ratio
between the urinary frusemide concentration and urinary sodium concentra-
Drug Interactions Involving Renal Excretory Mechanisms 179

tion. In nine subjects who received an intravenous infusion of frusemide and

inulin for over 2h, the administration of probenecid 500mg twice daily re-
duced the frusemide renal clearance to inulin clearance ratio from 0.92 ± 0.21
to 0.44 ± 0.18 (P < 0.05) (ANDREASEN et al. 1980). It was calculated that, when
frusemide was administered alone, 2 % of the urinary concentration of
frusemide was derived from filtration at the glomerulus with the remaining
98% being secreted by the organic anion transport system. When probenecid
was co-administered, the values were 8% and 92%, respectively. Probenecid
only slightly reduced urinary sodium excretion from 24.7% of the filtered load
to 21.0% (P < 0.05). These studies have supported the hypothesis that
frusemide's pharmacodynamic effects are more associated with the rate of
urinary frusemide excretion than with plasma concentrations, and have high-
lighted the usefulness of a competitive inhibitor of secretion to evaluate the
mechanisms of the natriuretic effect of this class of drug.

d) Indomethacin
BABER et al. (1978), in a study in 17 patients with rheumatoid arthritis, admin-
istered 25 mg indomethacin three times a day and demonstrated that
probenecid 0.5 g twice a day significantly increased plasma concentrations of
indomethacin by an average of 63%, but there was no difference (P> 0.05) in
the renal clearance of indomethacin (3.0 ± 2.5 mllh per kilogram indomethacin
alone versus 2.7 ± 2.5mllh per kilogram indomethacin plus probenecid). How-
ever, probenecid significantly (P < 0.01) reduced the apparent renal clearance
of the glucuronide conjugate of indomethacin from 271 ± 158 mllmin to 126 ±
188ml/min. This study highlighted a new type of renal drug interaction, in
which the renal clearance of a conjugated metabolite rather than the parent
drug was inhibited by another drug via competition for active tubular secre-
tion. However, it is not known whether the glucuronidation of indomethacin
occurs in the kidney (proximal tubule cells) and whether probenecid blocks
access of indomethacin to the site of renal glucuronidation or inhibits renal
glucuronidation.

e) Clofibrate
In four healthy male subjects who received clofibrate (500mg orally every
12h) with and without 6 hourly probenecid, steady-state concentrations of
unbound clofibric acid were increased almost fourfold (VEENENDAAL et al.
1981). The renal clearance of unbound clofibric acid was significantly (P <
0.05) reduced by probenecid from 14 ± 5 to 3 ± 1 ml/min, which is consistent
with inhibition of tubular secretion.

f) Sulphinpyrazone
In four patients with gout, the administration of probenecid reduced the renal
clearance of sulphinpyrazone from 21 ± 11 to 8 ± 5 ml/min, and when the
180 A. SOMOGYI

degree of protein binding (fu <0.02) was considered, the renal tubular secre-
tion of this organic anion was inhibited by a mean of 75% (as inulin clearance
remained unchanged) (PEREL et al. 1969).

g) Zidovudine
This antiviral agent undergoes extensive glucuronidation to form the ether
glucuronide GAZT (3'-azido-3' -deoxy-5'-~D-glucopyranuronosylthymidine)
(DE MIRANDA et al. 1989). In seven patients with AIDS or AIDS-related
complex, the administration of probenecid 500mg every 4h substantially al-
tered the disposition of oral zidovudine (2mg/kg 8 hourly). Probenecid signifi-
cantly (P < 0.01) increased the area underneath the plasma concentration-time
profile of zidovudine from 0.97 ± 0.04 to 2.00 ± O.77mg/1 x h. In addition, the
area underneath the plasma concentration-time curve of the glucuronide me-
tabolite was also significantly increased by probenecid from 6.9 ± 2.0 to 14.0 ±
5.7mg/1 x h (P < 0.01). In four of the subjects from whom urine could be
collected, the renal clearance of zidovudine was not altered by probenecid
(188 ± 65 vs. 141 ± 31mllmin per 1.73m2). However, the apparent renal
clearance of the glucuronide metabolite was substantially and significantly
reduced (P < 0.01) by probenecid from 293 ± 46 to 114 ± 50mllmin per 1.73m2
(Fig. 2). In a pilot study in two healthy subjects (HEDAYA et al. 1990),
probenecid 500mg every 6h for 3 days reduced the renal clearance of
zidovudine from 4.38 and 5.13mllmin per kilogram to 2.68 and 3.27mllmin per
kilogram, respectively. For the glucuronide conjugate, probenecid substan-
tially reduced the renal clearance from 11.6 and 11.0 to 2.82 and 2.44mllmin
per kilogram, respectively. Both groups attributed their findings to inhibition

5
--0-- GAZT PRE·PROBENECID
--0- GAZT POST·PROBENECID
4

o~~~~~-=~~~~~
o 2 4 6 8 10
Time (hours)
Fig. 2. Plasma concentration-time profile of the glucuronide conjugate of zidovudine
(GAZT) following administration of zidovudine alone (pre-probenecid) and after
concomitant administration of probenecid (post-probenecid) in seven patients.
Probenecid significantly reduced the apparent renal clearance of GAZT. (From DE
MIRANDA et al. 1989)
Drug Interactions Involving Renal Excretory Mechanisms 181

of the renal tubular secretion of this acidic glucuronide metabolite by

probenecid, in a similar manner to indomethacin (see Sect. C.I.1.d). However,
like the interaction with indomethacin, it is not known whether zidovudine
glucuronidation occurs in the kidney in humans and whether probenecid
prevents access of zidovudine to the proximal tubular cells (via competition
for transport into the cells) or whether it inhibits renal glucuronidation.

h) Acyclovir
In three patients, LASKIN et al. (1982) showed that probenecid 1 g orally 1 h
before acyclovir (administered as an intravenous infusion of 5mg/kg) signifi-
cantly (P < 0.05) reduced the renal clearance of acyclovir from 248 ± 80 to 168
± 48mllmin per 1.73m2 , with no alteration in creatinine clearance (a measure
of GFR). Acyclovir is both a weak acid (pKa 2.3) and weak base (pKa 9.3) and
it is likely, given the inhibition shown with probenecid, that the tubular secre-
tion of acyclovir is predominantly via the organic anion transporter.

i) Penicillins
Since the discovery that probenecid could increase serum penicillin concentra-
tions in humans (BURNELL and KIRBY 1951), many studies (too numerous to
cite) have shown that the renal clearance and hence tubular secretion of those
penicillins which are secreted can be reduced by probenecid, resulting in
elevated plasma concentrations. In a comprehensive study into the renal tubu-
lar secretion of benzylpenicillin and its inhibition by probenecid (OVERBOSCH
et al. 1988), four healthy volunteers underwent three different studies. In each
study, benzylpenicillin was infused to achieve three different steady-state
plasma concentrations. In the second and third studies, probenecid was admin-
istered by continuous infusion at low (31.3 mg/h in one subj ect, 21.4 mg/h in the
other three subjects) and high (between 93.8 and 200.3mg/h) doses resulting
in plasma probenecid concentrations of 6.7 and 35.1 mg/l, respectively.
Probenecid, in a concentration-dependent manner, increased the unbound
plasma concentrations of benzylpenicillin with an estimated probenecid ECso
of 52.3mg/1. When benzylpenicillin was administered alone at the three differ-
ent infusion rates, the data were able to be fitted to a non-linear equation in
which the maximum tubular excretion rate for benzylpenicillin (range 2493-
4131 mg/h) and ECso (plasma benzylpenicillin concentration at which half the
maximum excretion rate was achieved) of 22.6 to 60.5 mg/l could be derived.
When probenecid was administered in studies 2 and 3, the EDso (probenecid
dose at which 50% of the tubular transport is inhibited) was estimated to be
between 13.2 and 108.5 mg/h. These investigators extrapolated their data to
the clinical setting and suggested that the commonly used daily dose of 2 g/day
of probenecid is likely to be close to the maximum effective dose for inhibition
of the tubular secretion of benzylpenicillin.
Probenecid has also been shown to reduce the renal clearance of nafcillin
from 2.0 ± 0.5 mllmin per kilogram to 0.56 ± 0.17 mllmin per kilogram (P <
182 A. SOMOGYI

0.05) (WALLER et al. 1982) in five healthy subjects, and increase ticarcillin
plasma concentrations (CORVAIA et al. 1992); penicillins which have a renal
clearance greater than clearance due to glomerular filtration alone.
Of all the drugs interacting with probenecid, inhibition of the renal tubular
secretion of penicillins is considered to be the prototypic interaction for drugs
involving the kidney.

j) Quinolone Antibiotics
Several of these antibiotics are predominantly cleared by the kidney with renal
clearance values in excess of the glomerular filtration rate, indicating net
tubular secretion. In eight healthy male subjects, probenecid significantly (P <
0.01) reduced the renal clearance of cinoxacin to creatinine clearance ratio
from 1.32 ± 0.48 to 0.32 ± 0.17 (RODRIGUEZ et al. 1979). Since cinoxacin is
only 40% unbound in plasma, these data indicate that there is little tubular
reabsorption. Similar results were reported by ISRAEL et al. (1978), in which
probenecid reduced the renal clearance of cinoxacin from 153 ± 60mllmin per
1.73m2 to 66 ± 9mllmin per 1.73m2 (P < 0.001) in four healthy subjects. For
ciprofloxacin, with a renal clearance of over 200 ml/min, probenecid has been
shown to reduce its renal clearance by approximately 50% in healthy subjects
(WINGENDER et al. 1985); however, for fleroxacin, which has a renal clearance
of unbound drug of about 140ml/min, probenecid had no effect on its renal
clearance in five healthy subjects (WEIDEKAMM et al. 1987). These results again
support the hypothesis that for this class of antibiotic, if renal clearance is
greater than clearance by filtration, then tubular secretion is by, at least in part,
the organic anion transport system.

k) Cephalosporin Antibiotics
Interactions between cephalosporin antibiotics and probenecid have been
comprehensively reviewed by BROWN (1993). A large number of studies have
been published investigating the effect of probenecid on the disposition of the
cephalosporins. Some studies have shown a significant interaction to occur in
which plasma concentrations are increased, whilst others have not. The two
criteria for an interaction to occur in the kidney are, firstly, the kidney must
eliminate a large proportion of the drug. For the cephalosporins this can range
from 40% for ceftriaxone to 95% for cefuroxime (BROWN 1993). The second
criterion is that the renal clearance of the drug must exceed the clearance
attributed to filtration at the glomerulus. For some cephalosporin antibiotics
such as moxalactam, renal clearance does not exceed the glomerular filtration
clearance and hence probenecid has no effect on their renal clearance (Table
1). Two studies will be cited to illustrate the effects of probenecid on the renal
disposition of cephalosporins. In eight healthy male subjects, the effect of
probenecid 500 mg twice daily for 2 days prior to and on the morning of the
study was investigated to determine its effect on the disposition of cephradine
given as a 500-mg intravenous dose (ROBERTS et al. 1981). For cephradine,
Drug Interactions Involving Renal Excretory Mechanisms 183

Table 1. Association between evidence of net renal

tubular secretion of cephalosporin antibiotics and
inhibition of total body clearance by probenecid (Data
obtained from BROWN 1993; GILMAN et al. 1990). Note: the
list of cephalosporins is not exhaustive

Drug name Secreteda Secretion inhibited by

probenecidb

Cefaclor Yes Yes

Cefadroxil Yes Yes
Cephalexin Yes Yes
Cephalothin Yes Yes
Cefamandole Yes Yes
Cefazedone Yes Yes
Cefazolin Yes Yes
Cefmetazole Yes Yes
Cefmenoxime Yes Yes
Cefonicid Yes Yes
Ceforanide No No
Cefotaxime No No
Cefoxitin Yes Yes
Cefradine Yes Yes
Ceftazidine No No
Ceftizoxime No No
Ceftriaxone No No
Cefuroxime No No
Moxalactam No No

aResults of human pharmacokinetic studies indicate the

renal clearance of unbound drug is greater than GFR.
bTotal body clearance reduced by probenecid.

90% of the dose is excreted unchanged in urine, and in normal healthy subjects
renal clearance is approximately 400ml/min and the degree of plasma binding
is low. Hence this drug's major clearance is by tubular secretion. Probenecid
significantly (P < 0.01) increased the area under the serum cephradine concen-
tration-time curve from 24 ± 9 to 57 ± 11 mg/lx h. This was entirely due to a
significant (P < 0.01) reduction in renal clearance from a mean of 221/h to
12l/h. Interpretation of the data is consistent with inhibition of the renal
tubular secretion of cephradine by probenecid via the organic anion transport
system. As a second example, the disposition of cefonicid (500mg intramuscu-
larly) was investigated when probenecid (1 g orally) was administered to eight
healthy subjects (PITKIN et al. 1981). Approximately 90% of the dose of
cefonicid is excreted unchanged in urine, and its renal clearance is only ap-
proximately 30mllmin. However, this antibiotic is highly bound in plasma,
with an unbound fraction of 0.05. Hence although the renal clearance is low it
is mainly via tubular secretion. Probenecid significantly (P < 0.001) increased
the area under the cefonicid plasma concentration-time curve from 346 ±
34mg/1 x h to 724 ± 99mg/1 x h. The terminal half-life of cefonicid was
prolonged from 4.9 ± 0.8h to 7.5 ± 1.4h and the renal clearance of cefonicid
184 A. SOMOGYI

was substantially reduced in the 4- to 6-h time period from 30.2 ± 7.1 mllmin to
7.8 ± 2.0ml/min (P < 0.001) (PITKIN et al. 1981). The investigators calculated
that the secretion rate of cefonicid was reduced by probenecid from a mean of
744 to 246.ug/min. These data are again entirely consistent with inhibition
of renal tubular secretion. Table 1 is a summary of data on the effect of
probenecid on the renal disposition of the cephalosporin antibiotics. In all
cases, cephalosporin antibiotics whose disposition (hence renal clearance)
is altered by probenecid have renal clearance values indicating net tubular
secretion.

/) Famotidine
This organic cation histamine Hz antagonist has a renal clearance in humans of
almost three times the glomerular filtration rate, indicating net renal tubular
secretion presumably by the organic cation transport system. In eight young
healthy male volunteers (INOTSUME et al. 1990) the administration of
probenecid 1000mg 2h before, 250mg 1 h before and 250mg with a 20-mg oral
dose of famotidine resulted in the area underneath the plasma famotidine
concentration-time curve being significantly increased (P < 0.01) from 424 ±
54ng/ml x h to 768 ± HOng/ml x h. The renal clearance of famotidine was
reduced from 297 ± 54mllmin to 107 ± 14ml/min (P < 0.01). Using the
famotidine unbound fraction in plasma and creatinine clearance values, these
authors calculated that the clearance of famotidine due to tubular secretion
was significantly (P < 0.01) reduced by 89% (from 196 ± 61mllmin to 22 ±
12mllmin). The mechanism of this interaction is unclear as famotidine is an
organic cation, whereas probenecid's inhibitory effect is considered to be
specific for organic anions. It may well be that probenecid is a polysubstrate
which may interact with the organic anion transport system and suitable
substrates of the organic cation transport system. This has been investigated in
vitro with similar substances (ULLRICH et al. 1994).

2. Methotrexate
The antineoplastic and antirheumatoid drug methotrexate is approximately
50% excreted unchanged in urine (depending on the dose). Its renal clearance
is approximately 75 mllmin and the unbound fraction in plasma is 0.5. It has
been shown that methotrexate, being an organic anion undergoes renal tubu-
lar secretion by the organic anion transport system (LIEGLER et al. 1969) and
there is also some evidence for active tubular reabsorption (HENDEL and
NYFORS 1984). Although its renal clearance increases with urine pH, suggestive
of passive reabsorption, this was attributed to greater solubility of the drug in
urine (SAND and JACOBSEN 1981). Several drug interactions have been re-
ported involving the renal clearance of methotrexate. Since the drug has a
narrow therapeutic index, these interactions have sometimes been manifested
in overt methotrexate toxicity. Drugs which interact renally with methotrexate
are reviewed below.
Drug Interactions Involving Renal Excretory Mechanisms 185

a) Probenecid
In four patients administered intravenous methotrexate 200mg/m2 , the co-
administration of probenecid (500-1000mg) resulted in serum concentrations
at 24h which were four times higher (mean = OAmg/l) than with four other
patients not receiving probenecid (mean = 0.09mg/l) (AHERNE et a1. 1978). The
renal clearance of methotrexate was 104 ± 34ml/min, which was significantly
(P < 0.05) reduced to 46 ± 15 mllmin in those patients also taking probenecid.
In a more comprehensive study, three patients receiving high-dose metho-
trexate (3.0 g/m2) also received probenecid (dose not stated, but serum con-
centrations were between 21 and 116mg/l) (HOWELL et a1. 1979). There was a
significant increase in serum methotrexate concentrations, but, as only three
subjects were studied, the overall disposition of methotrexate was
not statistically significantly altered. Nevertheless, at 24h after the dose
of methotrexate, probenecid resulted in a threefold increase in plasma
methotrexate concentrations. In addition, in CSF at both 24 and 72h,
methotrexate concentrations were increased by a factor of 2.8- and 4.2-fold,
respectively. The half-life of methotrexate in CSF was not altered by
probenecid. This is a clinically important interaction as methotrexate toxicity
(bone marrow depression) has been reported in some patients receiving
probenecid. In the monkey, probenecid has been shown to completely inhibit
the renal tubular transport of methotrexate, resulting in renal clearance values
being equivalent to filtration at the glomerulus (BOURKE et a1. 1975).

b) Penicillins
There have been no detailed pharmacokinetic studies investigating the inter-
action between methotrexate and penicillins. However, a number of case
reports involving different penicillins have shown marked (penicillin by 35%,
piperacillin by 66%, ticarcillin by 40%, dicloxacillin by 93%) reductions in the
clearance of methotrexate (BLOOM et a1. 1986). The clinical importance of this
interaction was highlighted in patients receiving low-dose methotrexate for
psoriasis (MAYALL et a1. 1991). Five patients taking between 7.5 and 12.5mg
methotrexate daily developed neutropenia and thrombocytopenia when also
receiving amoxycillin, flucloxacillin, benzylpenicillin or piperacillin. Three of
these patients died. The clinical seriousness of this interaction highlights the
importance of drug interactions involving renal tubular secretion mechanisms.

c) Urinary Alkalinizers
Methotrexate can also undergo passive renal tubular reabsorption depending
on urine pH and the degree of ionization (SAND and JACOBSEN 1981). It has
been shown that alkalinization of urine (pH>7) results in a decrease in serum
methotrexate concentrations of between 40% and 77% (NIRENBERG et a1.
1977). Alkalinization of urine increases the degree of ionization of this weak
acid, preventing passive reabsorption and hence increasing its renal clearance.
However, SAND and JACOBSEN (1981) interpreted their findings in terms of
186 A. SOMOGYI

increasing the solubility of methotrexate in urine, although this was not di-
rectly assessed.

d) Non-steroidal Anti-inflammatory Drugs

There have been a number of pharmacokinetic studies and case reports high-
lighting a clinically and mechanistically significant interaction between
methotrexate and this class of drug.

a) Aspirin and Salicylates

Eight patients with rheumatoid arthritis and normal renal function were inves-
tigated to evaluate the effect of aspirin on the disposition of methotrexate
(STEWART et al. 1990a). On the 1st day of the study they were given lOmg
methotrexate by intravenous bolus and after 7 days of aspirin 975 mg orally
four times daily, another intravenous bolus dose of methotrexate was given.
Aspirin significantly (P < 0.05) increased the area underneath the plasma
methotrexate concentration-time curve from 2.7 ± 0.5 to 3.4 ± 0.9.umolll x h.
The total body clearance of methotrexate was reduced slightly from 141 ± 27
to 112 ± 23ml/min (P < 0.05) and the renal clearance of methotrexate was
reduced from 94 ± 25 to 77 ± 48mllmin; however, this was not statistically
significant. Due to the complex mechanism(s) (secretion and reabsorption) in
the renal clearance of methotrexate, collection of the 24-h urine in this study
may have been inappropriate to detect time- and concentration-dependent
changes in renal clearance. In nine patients receiving between 7.5 and 15mg
methotrexate orally, the addition of choline magnesium trisalicylate 2.25-4.5 g/
day decreased the total body clearance of methotrexate from 168 ± 50 to 128
± 57 ml/min (P < 0.05) and reduced the renal clearance of methotrexate from
117 ± 35 to 84 ± 27 ml/min (TRACY et al. 1992). In 12 rheumatoid arthritis
patients receiving low-dose (5-10 mg/m2 once a week) methotrexate, the effect
of aspirin (45-50mglkg per day for 16 days) on the intravenous pharmacoki-
netics of lOmglm2 methotrexate was investigated (FURST et al. 1990). Aspirin
had no effect on the total body clearance of methotrexate (urine was not
collected), or on the disposition of the metabolite 7-hydroxymethotrexate.
Finally, MAIER et al. (1986) proposed that the report by ST. CLAIR et al. (1985)
on methotrexate toxicity in three patients could have been attributed to the
high doses (3.9-5.0 glday) of aspirin received by those patients.

~) Indomethacin
There has been to systematic pharmacokinetic study investigating the interac-
tion with this nonsteroidal anti-inflammatory drug. However, there have been
case reports of acute renal failure and death, particularly in elderly patients
taking indomethacin and methotrexate (ELLISON and SERVI 1985). It is unclear
whether the mechanism involves competition between indomethacin and
methotrexate for active renal tubular secretion or, more likely, in the elderly
Drug Interactions Involving Renal Excretory Mechanisms 187

patient who requires the vasodilatory prostaglandins to maintain renal perfu-

sion, addition of indomethacin, which blocks the synthesis of these prostaglan-
dins, can result in acute renal failure resulting in reduced clearance of
methotrexate with its associated toxicities.

y) Ketoprofen
Although described as a pharmacokinetic study, details were lacking in a
retrospective study in 36 patients who were on a number of cycles of high-dose
(mean 3200mg/m2) methotrexate. Four of the nine patients who developed
severe toxicity were also taking ketoprofen and three of these patients died.
There was a substantial increase in the serum methotrexate concentration in
those patients with toxicity (THYSS et al. 1986).

8) Naproxen
In a comprehensive study, 15 patients aged between 30 and 78 years with
rheumatoid arthritis received oral and intravenous methotrexate 15 mg with
and without naproxen at a dose of 1000mg/day for 26 days (STEWART et al.
1990b). Twelve of the patients completed all four arms of the study and, for the
three who did not complete the study, withdrawal was not associated with drug
treatment. There was no significant difference in the total body clearance of
methotrexate with or without naproxen (103 ± 35 alone and 113 ± 48ml/min
with naproxen). In addition, the renal clearance of methotrexate from either
routes (oral and intravenous) was not altered by naproxen. It should be noted
that these patients had normal creatinine clearance values (80mllmin per
1.73m2). TRACY et al. (1992) showed significant (P < 0.05) but small decreases
in the total body clearance of methotrexate (168 ± 50-131 ± 42 ml/min) in nine
patients taking naproxen 750-1250mg/day, with no statistically significant
difference in renal clearance (117 ± 35 mllmin methotrexate alone and 96 ±
38ml/min with naproxen). AHERN et al. (1988) also found no difference in
plasma methotrexate concentrations in three patients with and without con-
comitant naproxen dosing. There has, however, been a report of serious tox-
icity in patients receiving methotrexate and naproxen (SINGH et al. 1986).

£) Ibuprofen
In nine patients with rheumatoid arthritis, the addition of ibuprofen 2.4-
3.6 g/day significantly (P < 0.05) reduced the total and renal clearances of
methotrexate from 168 ± 50 to 101 ± 37mllmin and from 117 ± 35 to 70 ± 27 mIl
min, respectively (TRACY et al. 1992).
Other non-steroidal anti-inflammatory drugs such as diclofenac, diflunisal,
flurbiprofen, sulindac and azapropazone have been investigated, albeit usually
with small numbers (n = 1-3) of patients. Hence, data interpretation has not
allowed conclusions to be deduced, although methotrexate toxicity occurred
in some patients. The interaction between methotrexate and non-steroidal
188 A. SOMOGYI

anti-inflammatory drugs is important as there have been several deaths related

to this interaction. The life-threatening nature of the interaction suggests that
non-steroidal anti-inflammatory drugs ought to be co-administered with ex-
treme caution to patients on methotrexate.
The mechanism for the interaction is likely to be one or a combination of:
(a) competition for active tubular secretion between the non-steroidal anti-
inflammatory drug and methotrexate as both are organic anions. Since
methotrexate renal clearance is concentration dependent and active secretion
appears to occur at low doses (HENDEL and NYFORS 1984), the interaction
could be more significant at low methotrexate doses; and (b) The non-steroi-
dal anti-inflammatory drugs can decrease renal perfusion via their inhibitory
action on cyclooxygenase, resulting in a reduction in renal vasodilatory pro-
staglandins, and a decrease in filtered and secreted methotrexate, causing
a reduction in renal clearance and increase in plasma concentrations. This
second mechanism for the interaction would be particularly relevant for
the elderly patient whose renal function can be easily compromised by non-
steroidal anti-inflammatory drugs.

II. Organic Cations

The most commonly investigated interacting drugs have been cimetidine,
other Hz receptor antagonists and trimethoprim.

1. Cimetidine
Cimetidine has been shown to reduce the renal clearance of the following
organic cations: procainamide and its active metabolite n-acetylprocainamide,
triamterene, ranitidine, amiloride, metformin, pindolol, nicotine and
quinidine.

a) Procainamideln-Acetylprocainamide
Several studies have been conducted in healthy subjects and patients and case
reports have highlighted the clinical importance of this interaction.
In the original study (SOMOGYI et al. 1983), six healthy subjects took 1 g
procainamide orally with and without 400mg cimetidine (1 h before the
procainamide dose and 200mg 4-hourly for 12h thereafter) with blood and
urine samples collected over 12 and 48h, respectively. Cimetidine significantly
(P < 0.025) increased the area under the plasma concentration-time curve of
procainamide by an average of 44% and prolonged the half-life from an
harmonic mean of 2.92 to 3.68h (Fig. 3). This was due to a marked reduction
(P < 0.025) in the renal clearance of procainamide from 347 ± 46 to 196 ± 11 mll
min (Fig. 4). In the same study, the area under the plasma concentration-time
curve for n-acetylprocainamide was significantly (P < 0.025) increased by an
average of 25% with a commensurate reduction (P < 0.025) in renal clearance
from 258 ± 60 to 197 ± 49mllmin. Similar data were obtained by CHRISTIAN et
Drug Interactions Involving Renal Excretory Mechanisms 189

100()

5·0

E
......
CI
~

--...
I:
.S! 1·0
CU

I:
CII
U
I:
0
U 0·5
CU
E/I)

.!
Q.

0·1 L.I.._....I.-_----L_.--l_ _.L.-_....I.-_--.J~

o 2 4 6 8 10 12
Time (h)

Fig. 3. Plasma concentration-time profiles of procainamide (round symbols) and

n-acetylprocainamide (square symbols) in six subjects who received 1 g oral
procainamide alone (open symbols) and when co-administered with cimetidine (closed
symbols). (From SOMOGYI et al. 1983)

al. (1984) in nine healthy subjects in which cimetidine increased the area under
the plasma concentration-time curve of procainamide by 40% (P < 0.0005),
prolonged its half-life by 24% and reduced its renal clearance by 36% (P <
0.0005). In addition, the area under the plasma concentration-time curve of n-
acetylprocainamide increased (P < 0.03) by 26% but there was no change in
renal clearance. In seven healthy Chinese subjects administered separate
single oral doses of 200mg and 400mg cimetidine (LAI et al. 1988), the area
under the plasma concentration-time curves of procainamide (500mg oral
dose) was increased by 24% (P < 0.01) and 38% (P < 0.05), respectively, with
significant (P < 0.01) reductions in renal clearance of 31 % and 41 %, respec-
tively. These data reinforced the mechanism of competitive inhibition of tubu-
190 A. SOMOGYI

10-12
Time Periods (h)

Fig. 4. Mean ± SD 2 hourly renal clearance values of procainamide when given alone
as 19 procainamide to six healthy subjects (open bars) and when co-administered with
cimetidine (hatched bars). The reduction in renal clearance by cimetidine was signifi-
cant (P < 0.02). (From SOMOGYI et al. 1983)

lar secretion of procainamide by cimetidine. Only the higher dose (400 mg) of
cimetidine significantly increased (mean 45%) the area under the plasma
concentration-time curve and reduced the renal clearance (16%) of n-acetyl-
procainamide. In a steady-state pharmacokinetic study in six healthy young
male subjects who received 500mg sustained release procainamide every 6h
for 13 doses, cimetidine 1200mg/day for 4 days significantly (P < 0.01) in-
creased the procainamide area under the plasma concentration-time curve by
a mean of 43% and significantly (P < 0.01) decreased renal clearance by a
mean of 34 % (RODVOLD et al. 1987). Steady-state serum n-acetylprocainamide
concentrations were significantly increased (P < 0.001) by a mean of 42%. The
results of these four studies in healthy subjects are similar and all groups
proposed the mechanism to involve inhibition of the active tubular secretions
of procainamide and n-acetylprocainamide by cimetidine via competitive
inhibition for the organic cation transporter. The interaction has been de-
scribed and mechanism verified in vitro (McKINNEY and SPEEG 1982). In a
study in 36 hospitalized male patients (aged 65-90 years) who were receiving
oral sustained release procainamide every 6 h, the addition of cimetidine
300mg every 6h for 3 days resulted in a significant (P < 0.005) increase in the
steady-state serum concentrations of procainamide (by a mean of 55%), and
n-acetylprocainamide (mean 36%) (BAUER et al. 1990). In ten of the patients
from whom urine was also collected, the ratio of pro cain amide renal clearance
to creatinine renal clearance was reduced by cimetidine by an average of 33 %
from 2.1 ± 0.6 to 1.4 ± 0.4 (P < 0.02). The n-acetylprocainamide to creatinine
clearance ratio also significantly (P < 0.05) declined from 1.4 ± 0.4 to 1.1 ± 0.4.
Twelve of these patients experienced mild to severe symptoms relating to
procainamide toxicity. Nine of the patients had symptoms of nausea, weak-
ness, malaise and PR interval increases of <20%. Three patients had severe
procainamide toxicity from prolongation of the PR interval of >25%, unifocal
premature ventricular contractions of <30/h and ventricular rates> 150/min for
Drug Interactions Involving Renal Excretory Mechanisms 191

greater than 30s. These latter patients had pro cain amide serum concentra-
tions of greater than 14mg/l and n-acetylprocainamide concentrations of
greater than 16mg/l. All patients with adverse effects had cimetidine discon-
tinued, and procainamide was also discontinued in those patients with severe
toxicity. The procainamide adverse effects disappeared within 24h after cessa-
tion of cimetidine therapy. In a case report (HIGBEE et al. 1984), a 71-year-old
male patient was being treated with procainamide (937.5mg every 6h) for
frequent multiform premature ventricular depolarizations and had achieved
steady-state therapeutic serum procainamide and n-acetylprocainamide con-
centrations of 9 and 7 mg/l, respectively. Following the addition of cimetidine
300mg every 6h for a benign gastric ulcer, the patient was re-admitted with
increased fatigue, weakness, nausea, anorexia and urinary retention. Examina-
tion revealed congestive heart failure and the ECG demonstrated sinus
rhythm with first-degree A V block accompanied by a decreased heart rate.
The procainamide and n-acetylprocainamide serum concentrations were both
15 mg/l, in the range associated with toxicity. The dose of procainamide was
subsequently decreased to 750mg every 6h and procainamide and n-
acetylprocainamide serum concentrations declined to 10 and 12mg/l, respec-
tively. Other dosage adjustments were made but the authors concluded by
stating that their case report should alert clinicians to the potentially danger-
ous adverse effects of this combination (cimetidine and procainamide) as well
as the need to evaluate carefully the source of gastrointestinal symptoms
which may be due either to ulcer disease or procainamide toxicity. They also
commented that the changes encountered suggested that older people may be
at a greater risk of pro cain amide toxicity when prescribed cimetidine.

b) Ranitidine
In a study in six male subjects, the administration of cimetidine 400mg 12
hourly resulted in a significant (P < 0.05) increase in the area under the plasma
ranitidine concentration-time curve from a single 150-mg oral dose of
ranitidine. This was associated with a significant reduction in the renal clear-
ance of ranitidine from 326 ± 67 to 244 ± 57 mllmin (VAN CRUGTEN et al. 1986).

c) Triamterene
In a similar design to the above study (MUIRHEAD et al. 1986), cimetidine
significantly (P < 0.03) increased the area underneath the plasma triamterene
concentration-time curve during a 24-h dosing interval of 100mg daily
triamterene. This was associated with a reduction in the renal clearance of
triamterene from 71 ± 46 to 21 ± 14ml/min. However, the decrease in renal
triamterene clearance occurred only during the first 6 h of the triamterene
dosing interval, at the time of maximum plasma cimetidine concentrations.
When the unbound fraction (0.4) of triamterene in plasma is taken into consid-
eration, the renal clearance due to tubular secretion (minus that due to
reabsorption) was reduced by cimetidine from a mean of 22mllmin to minus
192 A. SOMOGYI

28mllmin. The latter figure indicates that cimetidine completely blocked the
renal tubular secretion of triamterene and uncovered significant triamterene
passive reabsorption, a mechanism previously reported in vitro in the rat (KAu
1978). Cimetidine had no effect on the renal clearance of the major active
metabolite of triamterene, the sulphate conjugate of p-hydroxytriamterene
(MUIRHEAD et al. 1986).

d) Metformin
Cimetidine significantly (P < 0.01) increased the area under the plasma
metformin concentration-time curve by an average of 50% in seven young
healthy subjects given a 0.25-g daily dose of metformin and a 400-mg twice
daily dose of cimetidine. The increase in steady-state plasma metformin con-
centrations was associated with a significant (P < 0.01) reduction (mean 27 % )
in the renal clearance of metformin from 527 ± 165 ml/min to 378 ± 165 ml/min
(SOMOGYI et al. 1987). The reduction in metformin renal clearance only oc-
curred over the first 6h of cimetidine dosing, a phenomenon consistent with
competitive inhibition of metformin renal tubular secretion.

e) Pindolol
In a study to examine the potential stereoselectivity of the inhibition of renal
tubular secretion of organic bases by cimetidine, eight healthy young subjects
received a single 15-mg oral dose of racemic pindolol with 400mg cimetidine
taken twice daily for 2 days (SOMOGYI et al. 1992). The area under the plasma
concentration-time curve for R-(+) pindolol was significantly (P < 0.01) in-
creased by cimetidine from 230 ± 90 to 344 ± 78 ng/ml x h and for S-(-) pin dolo I
it was increased from 209 ± 73 to 288 ± 69ng/ml x h (P < 0.01). Cimetidine
significantly reduced the renal clearance of R-(+) pindolol from 170 ± 55mll
min to 104 ± 88ml/min (P < 0.01) and of S-(-) pindolol from 222 ± 66mllmin
to 155 ± 38mllmin (P < 0.01). The enantiomer with the higher renal clearance
[S-(-) pindolol] had a smaller (mean 26%) cimetidine-induced reduction
in renal clearance compared with R-(+) pindolol (mean 34%) (Fig. 5).
Cimetidine has a stereoselective inhibitory effect on the active transport of
organic cations in the proximal tubule.

f) Disopyramide
In a study in seven healthy male subjects, co-administration of cimetidine
400mg twice daily did not alter the disposition or the renal clearance of the
enantiomers of disopyramide when given as an intravenous bolus of 150mg
racemic disopyramide (BONDE et al. 1991). The reason for this lack of effect of
cimetidine on the renal clearance of a drug (an organic cation) which under-
goes extensive tubular secretion (renal clearance of unbound enantiomers of
disopyramide is 200-400mllmin) is not known. However, since the drug exhib-
its significant concentration-dependent plasma binding, the unbound concen-
trations need to be measured at all blood sampling times and not just at one
Drug Interactions Involving Renal Excretory Mechanisms 193

-s:::
E
:::;:
300

250 #
•
0
1m
R-(+)-pindolol
S-(-)-pindolol

-
R-(+)-pindolol + cimetidine
E
~ S-(-)-pindolol + cimetidine

-...
s::: 200
0
CD 150
u
CD
CJ) 100
...as
50
::s
.0
.-
::s 0

Fig. 5. Net renal clearance via tubular secretion for R-(+) pindolol and S-(-) pindolol
when racemic pindolol was administered alone (15 mg) and when co-administered with
cimetidine in eight subjects. Values represent mean ± SD. *P < 0.001, enantiomer alone
vs. enantiomer + cimetidine; #p < 0.001, R-(+) pindolol vs. S-(-) pindolol. (From
SOMOGYI et al. 1992)

particular time (or concentration) point. As was the case with triamterene and
metformin, the effect of cimetidine may only occur in the first few hours after
dosing, particularly with the low dose of cimetidine used_ Hence to observe an
effect on the renal clearance of disopyramide may have required frequent
urine collections and not reliance solely placed on urine collected over a long
time intervaL

g) Amiloride
In eight healthy subjects given a single dose of 5 mg amiloride, cimetidine
400 mg twice daily reduced the renal clearance of amiloride by an average of
17% from 358 ± 134 to 299 ± 118mllmin (P < 0.05), but there was no effect on
the area under the plasma amiloride concentration-time curve (SOMOGYI et aL
1989). The effect on renal amiloride clearance was small in magnitude and
occurred only in the 2- to 4-h time period after cimetidine dosing.

h) Nicotine
In six healthy subjects, cimetidine 600mg 12 hourly for 2 days significantly
(P < 0_05) reduced the renal clearance of nicotine (as an intravenous infusion
of 1,ug/kg per minute for 30min) from 1.49 ± 0.30 to 0.79 ± 0.30mllmin per
kilogram, whereas there was no effect (P> 0.05) on the renal clearance of the
nicotine metabolite cotinine (0.14 ± 0.04mllmin per kilogram without
cimetidine vs_ 0.15 ± 0.10mllmin per kilogram with cimetidine) (BENDAYAN et
al. 1990). Since the renal clearance of nicotine when co-administered with
cimetidine was below the renal clearance due solely to filtration (fu x GFR),
these results indicate the existence of passive reabsorption of nicotine in
addition to active tubular secretion.
194 A. SOMOGYI

i) Quinidine
In six healthy subjects, HARDY et al. (1983) showed that cimetidine given as
1.2g/day reduced the oral clearance of quinidine (as 400mg quinidine
sulphate) by 37% (P < 0.05) with no change in the urinary excretion of
unchanged quinidine. As a consequence, calculation of the renal clearance
allows one to conclude that cimetidine reduced the renal clearance of quini-
dine from 5.6 to 3.611h.

j) CephalothiniCephalexin
In order to investigate the specificity of cimetidine-induced reductions in the
renal tubular secretion of organic cations, VAN CRUGTEN et al. (1986) showed,
in six normal healthy subjects, that cimetidine (400mg twice daily) had no
effect (P> 0.05) on the renal clearance of the organic anion cephalothin (305
± 127mllmin alone vs. 284 ± 141mllmin with cimetidine), whose renal clear-
ance is inhibited by probenecid (Table 1). However, for the zwitterion cephal-
exin, cimetidine reduced (P < 0.01) the renal clearance from 263 ± 39 to 208 ±
21 mllmin. The renal clearance of this zwitterion, which has both acidic (pKa
2.6) and basic (pKa 7.0) functional groups, is also reduced by probenecid
(Table 1). These investigators concluded that cimetidine's effect was specific
for organic cations, and that, for those zwitterions that undergo tubular secre-
tion, both organic anion and cation transport systems might be involved.

2. Ranitidine
In the light of the inhibitory effect of cimetidine on the renal tubular secretion
of other organic cations, studies in humans have determined whether other Hz
antagonists (which are organic cations) behave in a similar manner to
cimetidine.

a) Procainamideln-Acetylprocainamide
In six healthy subjects, the administration of 150mg twice daily ranitidine l3h
prior to a 1-g oral procainamide dose increased the area under the plasma
procainamide concentration-time curve by an average of 14% with a signifi-
cant (P < 0.05) but small (mean 18%) reduction in the renal clearance of
procainamide (SOMOGYI and BOCHNER 1984). The area under the plasma n-
acetylprocainamide concentration-time curve was also reduced by an average
of 11 % (P < 0.05). In the same study, when the dose of ranitidine was increased
to 750mg in three subjects, the area under the plasma concentration-time
curve for procainamide was increased by an average of 20% and renal clear-
ance was reduced by an average of 35%, and for n-acetylprocainamide, renal
clearance was reduced by an average of 38%. This study demonstrated that in
the clinical doses used, ranitidine will have little inhibitory effect; however, if
larger doses are used or if the patient's renal function declines such that in
both cases higher plasma ranitidine concentrations are achieved, then inhibi-
Drug Interactions Involving Renal Excretory Mechanisms 195

tion of renal clearance of other organic cations can be observed. In another

study in six healthy male subjects who received 500mg sustained release
procainamide every 6h, the addition of ranitidine 150mg twice a day had no
effect on the plasma disposition or renal clearance of procainamide or n-
acetylprocainamide (RODVOLD et al. 1987). These results would be anticipated
in the light of the first study (SOMOGYI and BOCHNER 1984) and indicate that
any possible effect of ranitidine on the tubular secretion of procainamide or n-
acetylprocainamide may only occur at plasma ranitidine concentrations in
excess of those normally found following standard doses in patients with
normal renal function.

b) Triamterene
In eight healthy male subjects, ranitidine 150mg twice daily reduced the renal
clearance of triamterene (100mg daily for 4 days) by a mean of 51 % (P < 0.05)
from 94 ± 64 to 46 ± 42mllmin (MUIRHEAD et al. 1988). This was interpreted as
competition between triamterene and ranitidine for the renal organic cation
transport system. In the same study, the renal clearance of the sulphate conju-
gate of p-hydroxytriamterene was significantly (P < 0.01) reduced by ranitidine
from 105 ± 35 to 56 ± 20 mllmin. There was no effect of ranitidine on the degree
of binding of the sulphate conjugate to plasma components (fu = 0.09). Al-
though the results for triamterene are consistent with inhibition of tubular
secretion of triamterene by another organic cation (ranitidine), the inhibition
of the renal clearance of the sulphate conjugate raised the possibility of multi-
transport systems, as cimetidine which reduced triamterene renal clearance
did not alter the renal clearance of the sulphate conjugate (see Sect. C.IL1.c).

c) Nicotine
In six healthy subjects, ranitidine 300mg 12 hourly for 2 days had no effect on
the renal clearance of nicotine (intravenous infusion of 1 Jig/kg per minute for
30min) (1.49 ± 0.30 alone and 0.88 ± 0.40mllmin per kilogram with ranitidine)
or cotinine (BENDAYAN et al. 1990).

3. Famotidine

a) Procainamideln-Acetylprocainamide
Eight healthy subjects were given a single 5-mg/kg intravenous infusion of
procainamide alone and after 5 days of receiving 40mg famotidine once daily
(KLOTZ et al. 1985). Famotidine had no significant effect (P> 0.05) on the total
body clearance or renal clearance of procainamide and n-acetylprocainamide.
Although famotidine has a high renal clearance of more than 300ml/min, its
half-life is less than 3 h; hence plasma concentrations from a daily 40-mg dose
may have been insufficient to elicit inhibitory effects on the tubular secretion
of procainamide and n-acetylprocainamide. In addition, it has been observed
196 A. SOMOGYI

that the organic anion probenecid inhibits the tubular secretion of famotidine
(see Sect. c.I.l.l).

4. Trimethoprim

a) Procainamideln-Acetylprocainamide
In eight healthy subjects, procainamide 500mg orally every 6h for 3 days was
co-administered with and without trimethoprim 200mg daily for 4 days
(KOSOGLOU et al. 1988). Trimethoprim significantly (P < 0.05) increased the
steady-state plasma concentrations of procainamide by an average of 62 % and
for n-acetylprocainamide by 52%. These were due to significant (P < 0.05)
reductions in the renal clearances of procainamide (mean 47%) and n-
acetylprocainamide (mean 13%). These investigators detected small but
statistically significant (P < 0.0001) increases in the QTc interval when
trimethoprim was co-administered, presumably as a result of the elevated
plasma concentrations of procainamide and n-acetylprocainamide. The same
investigators also conducted a study in ten healthy young male subjects who
received a single l-g oral procainamide dose alone and with trimethoprim
100mg twice daily for 2 days prior to and 200mg with the procainamide dose
(VLASSES et al. 1989). The area under the plasma procainamide concentration-
time curve was significantly (P < 0.05) increased by trimethoprim by an aver-
age of 39% and the renal clearance was reduced from 487 ± 129 to 267 ±
123ml/min (P < 0.05). For n-acetylprocainamide, the area under the plasma
concentration-time curve was increased by an average of 27% by tri-
methoprim (P < 0.05) and renal clearance was reduced from 275 ± 78 to
192 ± 82mllmin (P < 0.05). Since trimethoprim is actively taken up by the
renal cortex via the organic cation transport system, and has a high renal
clearance (>220mllmin for the unbound drug), this interaction is consistent
with trimethoprim having a higher affinity than procainamide and n-
acetylprocainamide for the organic cation transporter and hence inhibiting
their tubular secretion.

b) Zidovudine
In nine HIV-infected patients with normal renal function, the effect of
trimethoprim (150mg orally) on the disposition of zidovudine (3mg/kg intra-
venous infusion) was evaluated (CHAlTON et al. 1992). Trimethoprim signifi-
cantly (P < 0.05) reduced the renal clearance of zidovudine from 0.34 ± 0.15 to
0.17 ± 0.091/h per kilogram, an average decrease of 58%. In addition, the
apparent renal clearance of the glucuronide conjugate of zidovudine was also
significantly (P < 0.05) reduced by trimethoprim from 0.70 ± 0.21 to 0.56 ±
0.1211h per kilogram. In the same study but on another occasion, the patients
received a single dose of trimethoprim 150 mg plus sulphamethoxazole 800 mg.
The decrease in zidovudine renal clearance averaged 48% and for the glucu-
ronide conjugate 20%, similar to when trimethoprim was administered alone,
Drug Interactions Involving Renal Excretory Mechanisms 197

indicating that the sulphonamide has no effect on zidovudine disposition.

Hence the renal tubular secretion of zidovudine is via the organic cation
transport system, whereas that of its glucuronide conjugate is by both organic
cation and anion systems as probenecid also reduced its renal clearance (see
Sect. c.I.l.g).

5. Amiodarone

a) Procainamideln-Acetyiprocainamide
The disposition of procainamide was investigated in eight patients who
received an intravenous dose of between 6 and 15 mg/kg before and after
1-2 weeks treatment with amiodarone (1.6g1day) (WINDLE et al. 1987).
Amiodarone significantly (P < 0.01) reduced the total body clearance of
procainamide from 0.43 ± 0.12 to 0.33 ± 0.12l/kg per hour. In two of the
patients, urine was also collected and amiodarone reduced the renal clearance
of pro cain amide from 0.25 and 0.241/kg per hour to 0.14 and 0.151/kg per hour,
respectively. Hence the reduction in total body clearance appears to be pre-
dominantly due to renal clearance, most likely a consequence of competition
for active tubular secretion. Since the half-life of amiodarone is almost 1
month, the degree of inhibition of procainamide renal tubular secretion might
be even higher when steady-state plasma concentrations of amiodarone are
reached after about 4 months of chronic dosing.

6. Quinine/Quinidine

a) Amantadine
In a study designed to evaluate the effect of age and gender on inhibition of
amantadine renal clearance by quinine and quinidine (GAUDRY et al. 1993), a
comparison was made between nine young 27- to 39-year-old subjects (five
male and four female) and nine 60- to 72-year-old adults (four male and five
female). Amantadine (an organic base undergoing tubular secretion) was
administered as a 3-mglkg oral syrup on the evening before the study and
quinine (as quinine sulphate 200mg) and quinidine (as quinidine sulphate
200mg) were administered orally approximately 12h later. Blood and urine
samples were collected at 2, 4, 6 and 8h after quinine/quinidine dosing and all
urine was collected in 2 hourly aliquots to 8h. Quinine and quinidine had no
significant (P> 0.05) effect on amantadine renal clearance in either young or
old subjects. However, when the data were separated according to gender,
quinine and quinidine significantly (P < 0.05) reduced the amantadine renal
clearance in males but not in females. When both male and female data were
combined, only quinine reduced the renal clearance of amantadine. The inves-
tigators cited animal studies which suggest that testosterone may enhance the
renal organic anion and cation transport of xenobiotics. However, it is unclear
as to how this would impact on inhibition of transport as there was no gender
198 A. SOMOGYI

difference in the renal clearance of amantadine. Clearly, further studies are

warranted to elucidate the mechanisms of this finding.

III. Organic Neutral Drugs

1. Digoxin
Digoxin is the most commonly used cardiac glycoside drug for the treatment of
congestive heart failure and atrial fibrillation. When administered by the intra-
venous route, approximately 60% of the dose is recovered unchanged in urine
and its renal clearance is in excess of the clearance via glomerular filtration
(taking into account the degree of plasma binding), indicating that this chemi-
cally neutral compound undergoes tubular secretion (STEINESS 1974). A large
number of interactions have been reported for digoxin and several of these
involve renal excretory mechanisms.

a) Amiodarone
Six patients with either ventricular or supraventricular tachyarrhythmias
were given a 1-mg intravenous dose of digoxin alone and after 2 weeks of
amiodarone 1600mg/day (NADEMANEE et al. 1984). The renal clearance of
digoxin was reduced by an average of 22% (from 0.85 ± 0.33 to 0.66 ± 0.39ml/
min per kilogram) but this was not significant (P> 0.05). However, in ten
healthy subjects who were administered 400 mg amiodarone daily for 3 weeks,
there was a mean 22% decrease (from 105 ± 39 to 84 ± 15mllmin; P < 0.05) in
the renal clearance of digoxin when administered as a 1-mg intravenous dose
(FENSTER et al. 1985). Amiodarone also reduced the non-renal clearance of
digoxin in both studies. The reduction in renal clearance was interpreted as a
reduction in the tubular secretion of digoxin as there was no effect on the renal
clearance of creatinine by amiodarone (NADEMANEE et al. 1984).

b) Spironolactone and Potassium-Sparing Diuretics

It was originally reported by STEINESS (1974) that spironolactone (100mg daily
for 10 days) decreased the renal clearance of digoxin by approximately 30%,
in nine patients stabilized on digoxin, to values almost identical to those for
inulin clearance. As there was no effect on the glomerular filtration rate, the
mechanism remained unexplained. The same group (WALDORFF et al. 1978)
also found that spironolactone (100mg twice daily for 5 days) given to four
hospitalized patients with arteriosclerotic heart disease and four healthy sub-
jects reduced the renal clearance of digoxin from 1.52 ± 0.86 to 1.13 ± 0.58mll
min per kilogram (P < 0.02). In a subsequent study in six healthy subjects,
spironolactone (100mglday for 8 days) was found to reduce the renal clear-
ance of digoxin from 168 ± 22 to 145 ± 17ml/min (P < 0.05), an average
decrease of 13% (HEDMAN et al. 1992). The mechanism for the interaction is
still unclear as it is not known by which active transport system in the kidney
Drug Interactions Involving Renal Excretory Mechanisms 199

digoxin is secreted (KOREN 1987). It has been reported that amiloride at a dose
of 5mg twice daily for 8 days in six healthy subjects increased (P < 0.001)
digoxin renal clearance from 1.3 ± 0.1 to 204 ± 0.5ml/min per kilogram follow-
ing a 15-.ug/kg intravenous bolus dose of digoxin without altering inulin clear-
ance (WALDORFF et al. 1981). The mechanism by which amiloride increases the
renal clearance of digoxin was attributed to increased potassium in distal
tubular cells and therefore increasing digoxin secretion, although this latter
mechanism was not proven.

c) Quinidine
Several mechanisms, including a reduction in digoxin renal clearance, are
involved in the interaction of quinidine on digoxin disposition. The reduc-
tion in the renal clearance of digoxin by quinidine has been reported as: (a) a
mean of 35% reduction in three patients on long-term oral digoxin and admin-
istered 250mg twice daily quinidine, with no effect on creatinine clearance
(HOOYMANS and MERKUS 1978); (b) a mean of 33% reduction (from 1.64 ± 0.60
to 1.09 ± 0.24mllmin per kilogram; P < 0.05) in six patients given a 1-mg
intravenous bolus dose of digoxin with and without 200mg quinidine 6 hourly
for 7 days (HAGER et al. 1979); (c) a mean of 34% reduction (from 53 ± 21 to
35 ± 13mllmin per 1.73m2 ; P < 0.001) in 15 patients prescribed between 0.125
and 0.25mg/day digoxin and 200-300mg/day quinidine, with no alteration in
creatinine clearance (MUNGALL et al. 1980); (d) a mean of 52% decrease (from
130 ± 32 to 62 ± 10ml/min; P < 0.01) in digoxin renal clearance in five patients
given an intravenous dose of tritiated digoxin with and without quinidine 0.6 g
twice daily (SCHENCK-GUSTAFSSON and DAHLQVIST 1981); (e) a decrease of
32% in mean renal clearance of digoxin (from 1.93 ± 0.61 to 1.32 ± Oo4Oml/min
per kilogram) in seven healthy subjects receiving quinidine 800mg/day and
increasing to a mean 54% reduction in digoxin renal clearance (from 1.47 ±
0.30 to 0.68 ± 0.22; P < 0.001) when the dose of quinidine was doubled to
1600mg/day for 4 days (LEAHEY et al. 1981). Digoxin was administered as a
1-mg intravenous infusion over 60min on both occasions in both groups. A
relationship (r = 0.60) was found between the serum quinidine concentration
and reduction in digoxin renal clearance. Thus, the reduction in renal clear-
ance of digoxin depends on the concentration of quinidine and is consistent
with competition for active tubular secretion; (f) In a comprehensive study in
eight healthy young male subjects, digoxin 0.5-0.75mglday was given alone
and with 400mg twice daily quinidine (in four subjects). The renal clearance of
digoxin was significantly reduced (P < 0.05) by an average of 29% from 155 ±
40 to 110 ± 21 mllmin. In the other four subjects, the effect of quinine (the
diastereoisomer of quinidine) given as 750 mg/day was investigated. The renal
clearance of digoxin was not altered by quinine (177 ± 40 ml/min digoxin alone
and 185 ± 53 ml/min with quinine) (HEDMAN et al. 1990).
Since quinidine has no effect on glomerular filtration rate, the results of
these studies have been interpreted as a reduction in the tubular secretion of
200 A. SOMOGYI

digoxin by quinidine. It is possible that both digoxin and quinidine compete for
the same carrier-mediated transport system in the kidney. Quinidine competes
for the organic cation transporter; however, it has been shown in animals that
digoxin is not eliminated by this transport system (KOREN 1987). Clearly
further studies are warranted to delineate the mechanism of the renal tubular
secretion of digoxin.

d) Verapamil
Verapamil has been shown to increase plasma digoxin concentrations and
produce toxicity and in some cases fatality (as reviewed by STOCKLEY 1994). In
eight healthy subjects who received a l-mg intravenous infusion of digoxin
over 15 min, the addition of verapamil 800mg orally three times daily for 10
days resulted in a reduction in the renal clearance of digoxin from 2.2 ± 0.4 to
1.7 ± 0.5mllmin per kilogram (P < 0.025) (PEDERSEN et al. 1981). Similarly, in
seven patients receiving verapamil (80mg three times daily) for long-term
therapy, the renal clearance of digoxin from a dose of 125 J.lglday was reduced
after 1 week of verapamil therapy by an average of 36% (from 198 ± 46 to 128
± 27ml/min; P < 0.001), however, after 6 weeks of verapamil there was no
reduction in the renal clearance of digoxin (203 ± 71 mllmin) (PEDERSEN et al.
1982). In a study by BELZ et al. (1983) in healthy subjects who were adminis-
tered digoxin 0.375mglday to steady state, the addition of verapamil (80mg
three times/day for 2 weeks) reduced (P < 0.05) the digoxin renal clearance
from 218 ± 90 to 148 ± 42mllmin per 1.73m2 • In a more recent study in six
patients with chronic atrial fibrillation, verapamil (240mg/day) had no effect
on the renal clearance of digoxin (dose varied between 0.25 and O.5mg/day)
(HEDMAN et al. 1991). The renal clearance of digoxin was 153 ± 51mllmin
when given alone and 173 ± 51 ml/min when given with verapamil (P > 0.05).
This study found a significant association between the renal clearance of
digoxin and urine flow rate (r = 0.75; P < 0.001). At a urine flow rate of lmll
min the renal clearance of digoxin was about 100mllmin and at 4mllmin the
renal clearance had doubled to almost 200mllmin. There was no difference in
urine flow rate between the two study occasions. It would appear that the
effect of verapamil on digoxin renal clearance may be transient and may be
confounded by alterations in urine flow rate.

e) Other Calcium Antagonists

There have been several studies investigating the effect of other calcium
antagonists (diltiazem, nifedipine) on the disposition of digoxin. Most of the
studies have resulted in contradictory findings and the magnitude of any
change in the disposition of digoxin has usually been small. Effects on the
renal clearance of digoxin have usually not been investigated, but based on
effects on the total clearance of digoxin (when they occur) it could be con-
cluded that these other calcium antagonists do not alter digoxin renal clear-
ance (RODIN and JOHNSON 1988).
Drug Interactions Involving Renal Excretory Mechanisms 201

The mechanism for the renal tubular secretion of digoxin remains un-
known. Studies in animals have indicated that digoxin is not secreted by the
organic cation or anion transport system. There was also no evidence to
indicate that binding to the antiluminal Na+/K+-ATPase alters the renal clear-
ance of digoxin (KOREN 1987).

D. Tubular Reabsorption

I. Proximal Tubule Site

1. Lithium
Lithium is used for the management of manic depressive psychoses, and has a
relatively narrow therapeutic index in which toxicity is manifested by impaired
consciousness, disorientation, ataxia, tremor, muscle twitches and vomiting.
These adverse effects usually occur at plasma concentrations greater than
2mmolll. The kidney is the major clearance organ for the elimination of
lithium. After filtration of lithium and sodium at the glomerulus, approxi-
mately 70% of the filtered load of sodium and lithium is reabsorbed at the
proximal tubule, which does not distinguish between these two inorganic ions.
Thus, the renal clearance of lithium is similar and parallels that of sodium and
the ratio of the renal clearance of lithium to that of creatinine is approximately
0.2. A number of drugs alter the clearance of lithium through altering its renal
clearance. These include non-steroidal anti-inflammatory drugs and thiazide
diuretics, which reduce the renal clearance of lithium, and sodium salts and
theophylline, which increase the renal clearance of lithium.

a) Theophylline
In ten healthy subjects with normal renal function administered lithium car-
bonate 900mg daily for 7 days, the addition of theophylline to achieve steady-
state plasma concentrations of between 5 and 12mg/1 significantly (P < 0.05)
increased the renal clearance of lithium by an average of 42% from 18.3 ± 7.0
to 26.1 ± 11.5mllmin and shortened the half-life from 28.1 ± 7.5 to 21.8 ± 6.6h
(PERRY et al. 1984). There was also a significant decrease in serum lithium
concentrations by an average of 30%. This interaction was dependent on the
plasma theophylline concentration, as there was a statistically significant (P <
0.05) linear relationship between the plasma theophylline concentration and
the percentage change in lithium clearance. The incidence of side effects
(restlessness, tremor and anorexia) has been reported to be significantly
greater when theophylline was co-administered to patients (COOK et al. 1985).
In a single case report of a patient with mania treated with lithium, the
addition of theophylline resulted in a rapid onset of relapse of mania. When
the dosage of lithium was increased, serum lithium concentrations increased
202 A. SOMOGYI

and the patient's mania was controlled (SIERLES and OSSOWSKI 1982). The
mechanism for the increase in the renal clearance of lithium by theophylline is
unknown. Theophylline does not alter renal blood flow or glomerular filtration
rate and therefore appears to reduce the tubular reabsorption of lithium.

b) Sodium Salts
THOMSEN and SCHOU (1968) originally reported that administration of sodium
bicarbonate with lithium resulted in a 30% increase in the ratio of lithium
renal clearance to creatinine clearance, and DEMERS and HENINGER (1971)
reported in six patients with mania, receiving lithium 1200-1800mg/day, that
when a high-sodium diet (120-250mmol/day) was compared to a low-sodium
(69-91 mmoI/day) diet, there was a decline in the serum lithium concentration
from a mean of 0.99 to 0.83 mmol/l, and an increase in manic effects and
behaviour ratings. In a case report, a male with depressive illness was pre-
scribed 250 mg lithium carbonate four times daily to achieve a target lithium
serum concentration of 0.5 mmol/I. When the frequency of dosage was in-
creased to six times a day his serum lithium concentration was still below
0.6mmoI/1 as the patient was also taking sodium bicarbonate. When this was
ceased, serum lithium concentrations reached 0.8mmol/1 on the initial dosage
(ARTHUR 1975). In a subsequent case study, MCSWIGGAN (1978) reported on
several patients who failed to achieve and maintain therapeutic serum lithium
concentrations. It was subsequently revealed that a nurse had been adminis-
tering a proprietary saline drink for upset stomachs, which contained about
50% sodium bicarbonate. When this was ceased, two patients developed
lithium toxicity. Other studies have shown similar effects in that intake of
sodium salts reduces serum lithium concentrations, resulting in a reduced
pharmacodynamic effect of lithium (DEMERS and HENINGER 1971).
The mechanism for this interaction is consistent with the known mecha-
nism for the renal elimination of lithium. When plasma sodium concentrations
are elevated through the addition of sodium-containing preparations, extracel-
lular fluid is expanded and the kidney reduces its reabsorption of sodium and,
as a consequence, lithium, resulting in increased renal clearance and reduced
plasma concentrations. This interaction is clinically important and has implica-
tions for the management of mania in patients.

c) Diuretics

a) Thiazide Diuretics
There has been no systematic pharmacokinetic study investigating the interac-
tion between lithium and thiazide diuretics. Nevertheless, there have been a
number of clinically important case reports, which have highlighted a poten-
tially serious interaction between these drugs in which thiazide diuretics in-
crease serum lithium concentrations. For example, in 22 patients receiving
hydroflumethiazide or bendrofluazide, the renal clearance of lithium was sig-
Drug Interactions Involving Renal Excretory Mechanisms 203

nificantly (P < O.OS) reduced from a mean of 20ml/min to IS ml/min (PETERSEN

et al. 1974) and in a patient treated with lithium carbonate, addition of SOOmg
chlorothiazide daily raised the serum lithium concentration from 1.3 to
2.0mmol/l (LEVY et al. 1973). HIMMELHOCH et al. (1977) showed an association
between the percentage reduction in calculated lithium renal clearance and
the dose of chlorothiazide (SOOmg/day, a 40% decrease, 1000mg/day, a 68%
decrease). KERRY et al. (1980) reported on a patient stabilized on lithium
(serum concentration 0.9-1.2mmol/l) in which the addition of bendrofluazide
increased the serum lithium concentration to 2.4 mmol/l and in four normal
healthy subjects administered SOmg/day hydrochlorothiazide there was a sig-
nificant (P < 0.01) increase in serum lithium concentrations from 0.4 to
0.S3mmol/l (JEFFERSON and KALIN 1979).
The mechanism for this interaction is still not completely understood.
However, the interaction would appear to be of a long-term nature rather than
short term, as in a study in healthy subjects a single dose of bendrofluazide had
no effect on the renal excretion of lithium (THOMSEN and SCHOU 1968). It has
been proposed (STOCKLEY 1994) that thiazide diuretics, although initially elic-
iting an increase in sodium excretion through their major effect in the distal
tubule, upon chronic administration the proximal tubule compensates for the
loss of sodium by retaining sodium. As both sodium and lithium are co-
reabsorbed, lithium would also be retained within the body, a result of the
reduction in renal clearance and therefore increased plasma concentrations.

~) Loop Diuretics
There have been no comprehensive studies investigating potential interactions
between loop diuretics and lithium, although there have been a number of case
reports and some studies in normal healthy subjects. In five normal healthy
subjects administered 300mg lithium three times daily, resulting in steady-
state serum concentrations of 0.4 mmolll, addition of 40 mg frusemide daily for
14 days produced no lithium-induced adverse effects or increase in serum
lithium concentrations in five of six subjects. However, one subject withdrew
from the study because of toxicity which was associated with a 61 % increase
in serum lithium concentration (from 0.44 to 0.71 mmol/l; JEFFERSON and
KALIN 1979). In a case report bumetanide has also been associated with the
development of lithium toxicity in a patient (KERRY et al. 1980), in which the
serum lithium concentration increased from between 0.8 and 1.1mmol/l to
1.6mmol/l.
The mechanism for this interaction is likely to be similar to that described
for the thiazide diuretics.

d) Non-steroidal Anti-inflammatory Drugs

Several clinical studies and case reports have demonstrated increases in serum
lithium concentrations with the use of various but not all non-steroidal anti-
inflammatory drugs.
204 A. SOMOGYI

a) Indomethacin
Although published in abstract from (LEFTWICH et al. 1978), it was reported
that in five healthy subjects taking 300-600mg lithium carbonate daily, addi-
tion of indomethacin 50mg three times a day significantly reduced the renal
clearance of lithium by an average of 31 % with a significant (P < 0.05) 43%
increase in serum lithium concentrations. The urinary excretion of the me-
tabolite of prostaglandin (PGE z) synthesis was reduced by indomethacin. In
ten normal healthy female subjects on a balanced sodium diet, addition of
150mg indomethacin daily to lithium sulphate (330mg twice daily) signifi-
cantly increased the serum lithium concentration from 0.68 ± 0.07 mmol/l to
0.84 ± O.13mmol/l (REIMANN et al. 1983). This was accompanied by a signifi-
cant decrease in the renal clearance of lithium (from 26 ± 7 mIlmin to 19 ± 5 mIl
min; P < 0.001) and urinary PGE z excretion was reduced from 298 ± 46 to 168
± 89ng/24h (P < 0.005). Reports of lithium toxicity associated with increased
serum lithium concentrations in patients also taking indomethacin have been
described (RAGHEB et al. 1980).

~) Ibuprofen
In a study in 11 healthy young (23-38 years) subjects, ibuprofen 1600mg daily
increased the serum lithium concentration (resulting from a dose of 450mg
lithium sulphate twice daily) by an average of 15%, from 0.67 ± 0.12 to 0.77 ±
O.13mmoIlI (P < 0.001) (KRISTOFF et al. 1986), with a significant decrease in
renal clearance (24 ± 5 to 19 ± 4mIlmin, n = 5). In patients maintained on
lithium, the addition of 1800mg ibuprofen increased serum lithium concentra-
tions by between 12% and 66% (RAGHEB 1987a). There have been reports of
lithium toxicity in patients receiving ibuprofen (KHAN 1991).

y) Diclofenac
Five young healthy women, on a standardized sodium diet of 150mmoIlday,
were administered lithium as a sustained release tablet of 330mg twice daily.
Diclofenac 50mg three times daily was also administered (on another occa-
sion) and it significantly (P = 0.002) decreased the renal lithium clearance by an
average of 23% and significantly (P < 0.001) increased serum lithium concen-
trations by a mean of26%. Renal prostaglandin synthesis was measured by the
urinary excretion of PGEz and was decreased by 53% (mean) compared with
the control period (213 ± 81 to 103 ± 36 ng/24 h) (REIMANN and FROLICH 1981).

0) Naproxen
In seven patients stabilized on lithium, the addition of 750mg naproxen re-
sulted in a 16% (range 0%-42%) increase in serum lithium concentrations
(from 0.81 ± 0.11 to 0.94 ± 0.08mmol/l). The renal clearance of lithium de-
creased from 17 ± 5 to 13 ± 4mllmin (P < 0.05) (RAGHEB and POWELL 1986).
One patient developed signs of toxicity and had an increase in the serum
lithium concentration from 0.95 to 1.13 mmoIll.
Drug Interactions Involving Renal Excretory Mechanisms 205

€) Piroxicam
A case report documented an increase in serum lithium concentrations in a
patient also taking piroxicam which resulted in adverse effects (KERRY et al.
1983).

~) Phenylbutazone
Although no systematic pharmacokinetic study has been conducted, there has
been a case report in which, in six patients with a bipolar affective disorder,
treatment with 300mg phenylbutazone caused a small increase (range 0%-
15%) in serum lithium concentrations, which was associated in some patients
with side effects (RAGHEB 1990).

11) Sulindac
In six patients requiring lithium therapy at doses ranging from 600 to 900mgl
day, the addition of sulindac 150mg twice a day did not significantly (P> 0.05)
alter serum lithium concentrations (0.84 ± 0.07 lithium alone vs. 0.83 ±
O.lOmmolll with sulindac) (RAGHEB and POWELL 1986).

e) Aspirin and Salicylate

Six healthy female subjects aged 22-46 years were placed on a strict sodium
diet of 150mmol/day (REIMANN et al. 1985). Lithium was taken as sustained
release tablets of 330 mg twice a day to achieve steady-state plasma concentra-
tions of between 0.6 and 0.8mmol/l. The subjects received aspirin as an intra-
venous loading dose of 0.5g followed by 8.8mg/min for 170 min. Urinary
prostaglandin excretion as measured by PGE 2 was significantly reduced by
aspirin from 155 ± 95 to 61 ± 28ng per 3h (P < 0.05). However, there was no
significant effect of aspirin on renal blood flow, glomerular filtration rate,
plasma lithium concentration or renal clearance of lithium (38 ± 11 to 35 ±
12ml/min; P> 0.05). In the same study, sodium salicylate (1.8g over 3h) had
no effect on urinary prostaglandin excretion or serum lithium concentration,
but significantly reduced the renal clearance of lithium to 29 ± 5 mllmin (P <
0.05). In seven patients stabilized on lithium, the addition of 3.9 g aspirin daily
had no effect on serum lithium concentrations or renal lithium clearance
(RAGHEB 1987b), and in ten normal female subjects, aspirin (1 g four times
daily) had no effect on serum lithium concentrations or renal lithium clearance
but did reduce renal prostaglandin excretion by between 65% and 70%
(REIMANN et al. 1983).
The mechanism for the interaction between non-steroidal anti-inflamma-
tory drugs and lithium has been attributed to inhibition of the synthesis of
renal vasodilatory prostaglandins (which has been documented), resulting in a
reduction in renal blood flow and glomerular filtration rate and as a conse-
quence, reduction in the renal clearance of lithium. However, it is unclear as to
why aspirin has no effect on renal lithium clearance but inhibits renal prosta-
206 A. SOMOGYI

glandin synthesis. The renal sparing effect of sulindac would be consistent with
its lack of interaction with lithium. This interaction is clinically important as a
number of reports of lithium toxicity due to non-steroidal anti-inflammatory
drugs excluding sulindac and aspirin/salicylate have been documented.

II. Distal Tubule/Collecting Duct Site

1. Urine pH and Flow Rate

Many studies have been conducted to examine the influence of urine pH or
flow rate on the renal clearance of drugs via modifying passive reabsorption
(see Sect. B.lV). Two examples will be presented, one for organic anions
(chlorpropamide) and one for organic cations (pseudoephedrine).

a) Chlorpropamide
Six healthy subjects received a 250-mg oral dose of chlorpropamide on
several occasions, once when urine pH was made acidic (pH 4.7-5.5) with
ammonium chloride and once when alkaline (pH 7.1-8.2) with sodium bicar-
bonate (NEUVONEN and KARKKAINEN 1983). The area under the plasma
chlorpropamide concentration-time curve was significantly (P < 0.001) larger
when urine was acidic (3635 ± 1212mg/1 x h) than when alkaline (700 ± 98mg/
I x h), and there were clear differences in half-life from a mean of 13h under
alkaline urine conditions to 69 h when urine was acidic. These substantial
differences were attributed to a marked dependence of renal clearance on
urine pH. When all the data were combined, providing 144 sets of urine pH
and renal clearance data, it was noted that, at a urine pH of 5, renal clearance
was 1 mllh but, when urine pH was 7, renal clearance was 100mllh, a two
orders of magnitude difference. This is probably the best example of how urine
pH can influence the disposition of a drug. For chlorpropamide, an acidic,
lipophilic drug with a pKa of 4.8, by raising urine pH, the degree of ionization
is increased thus preventing passive reabsorption and increasing apparent
renal clearance; in contrast, when urine pH is acidic, the degree of unioniza-
tion is increased, resulting in increased passive reabsorption and negligible
renal clearance (0.017 mllmin).

b) Pseudoephedrine
In eight subjects, who received a 5-mg/kg oral dose of pseudoephedrine,
BRATER et al. (1980) found a strong relationship between the half-life of
pseudoephedrine (a surrogate marker of renal clearance) and urine pH, such
that at pH 5.8 the half-life was about 300min and at pH 7.2 it was 600min.
These investigators also found that for each subject, when urine was alkaline
(pH >7), there was a significant correlation (r = 0.71-0.99) between renal
clearance and urine flow. Thus for this lipophilic organic cation with a pKa of
9.4, its renal clearance is highly dependent on both urine pH and flow rate. The
Drug Interactions Involving Renal Excretory Mechanisms 207

observations found were in agreement with those predicted, thus confirming

the mechanism of passive reabsorption and its modulation by the physiological
variables of urine pH and flow rate.

Acknowledgement. This review has been supported by the National Health and Medi-
cal Research Council of Australia.

References
Ahern M, Booth J, Loxton A, McCarthy P, Meffin P, Kevat S (1988) Methotrexate
kinetics in rheumatoid arthritis: is there an interaction with nonsteroidal anti-
inflammatory drugs? J RheumatoI15:1356-1360
Aherne GW, Piall E, Marks V, Mould G, White WF (1978) Prolongation and enhance-
ment of serum methotrexate concentrations by probenecid. Br Med J 1:1097-1099
Andreasen F, Sigurd B, Steiness E (1980) Effect of probenecid on excretion and
natriuretic action of furosemide. Eur J Clin Pharmacol 18:489--495
Arthur RK (1975) Lithium levels and "Soda BIC". Med J Aust 2:918
Baber N, Halliday L, Sibeon R, Littler T, Orme M L'E (1978) The interaction between
indomethacin and probenecid. Clin Pharmacol Ther 24:298-307
Bauer LA, Black D, Gensler A (1990) Procainamide-cimetidine drug interaction in
elderly male patients. J Am Geriatr Soc 38:467-469
Belz GG, Doering W, Munkes R, Matthews J (1983) Interaction between digoxin and
calcium antagonists and antiarrhythmic drugs. Clin Pharmacol Ther 33:410-417
Bendayan R, Sullivan JT, Shaw C, Frecker RC, Sellers EM (1990) Effect of cimetidine
and ranitidine on the hepatic and renal elimination of nicotine in humans. Eur J
Clin PharmacoI38:165-169
Beyer KJ, Russo HF, Tillson EK, Miller AK, Verwey WF, Gass SR (1951) "Benemid",
p-(di-n-propylsulfamyl)-benzoic acid: its renal affinity and its elimination. Am J
Physiol 166:625-640
Bloom EJ, Ignoffo RJ, Reis CA, Cadman E (1986) Delayed clearance (CL) of
methotrexate (MTX) associated with antibiotics and antiinflammatory agents.
Clin Res 34:560A
Bonde J, Pedersen LE, Nygaard E, Ramsing T, Angelo HR, Kampmann JP (1991)
Stereos elective pharmacokinetics of disopyramide and interaction with
cimetidine. Br J Clin Pharmacol 31:708-710
Borga 0, Larsson R, Lunell E (1986) Effects of probenecid on enprofylline kinetics in
man. Eur J Clin Pharmacol 30:221-223
Bourke RS, Cheda G, Bremer A, Watanabe 0, Tower DB (1975) Inhibition of renal
tubular transport of methotrexate by probenecid. Cancer Res 35:110-116
Brater DC, Kaojarern S, Benet LZ, Lin ET, Lockwood T, Morris RC, McSherry EJ,
Melmon KL (1980) Renal excretion of pseudoephedrine. Clin Pharmacol Ther
28:690-694
Brown GR (1993) Cephalosporin-probenecid drug interactions. Clin Pharmacokinet
24:289-300
Burnell JM, Kirby WMM (1951) Effectiveness of a new compound, benemid, in elevat-
ing serum penicillin concentrations. J Clin Invest 30:697-700
Chatton JY, Munafo A, Chave JP, Steinhauslin F, Roch-Rame1 F, Glauser MP, Biollaz
J (1992) Trimethoprim, alone or in combination with sulphamethoxazole, de-
creases the renal tubular secretion of zidovudine and its glucuronide. Br J Clin
PharmacoI34:551-554
Christian CD, Meredith CG, Speeg KV (1984) Cimetidine inhibits renal pro cain amide
clearance. Clin Pharmacol Ther 36:221-227
Cook BL, Smith RE, Perry PJ, Calloway RA (1985) Theophylline-lithium interaction.
J Clin Psychiatry 46:278-279
208 A. SOMOGYI

Corvaia L, Li SC, Ioannides-Demos LL, Bowes G, Spicer WJ, Spelman DW, Tong N,
McLean AJ (1992) A prospective study of the effects of oral probenecid on the
pharmacokinetics of intravenous ticarcillin in patients with cystic fibrosis. J
Antimicrob Chemother 30:875-878
De Miranda P, Good SS, Yarchoan R, Thomas RV, Blum MR, Myers CE, Broder S
(1989) Alteration of zidovudine pharmacokinetics by probenecid in patients with
AIDS or AIDS-related complex. Clin Pharmacol Ther 46:494-500
Demers RG, Heninger GR (1971) Sodium intake and lithium treatment in mania. Am
J Psychiatry 128:132-136
Elion GB, Yii T-F, Gutman AB, Hitchings GH (1968) Renal clearance of oxipurinol,
the chief metabolite of allopurinol. Am J Med 45:69-77
Ellison NM, Servi RJ (1985) Acute renal failure and death following sequential inter-
mediate-dose methotrexate and 5-FU: a possible adverse effect due to concomi-
tant indomethacin administration. Cancer Treat Rep 69:342-343
Fenster PE, White NW, Hanson CD (1985) Pharmacokinetic evaluation of the digoxin-
amiodarone interaction. J Am Coll Cardiol5:108-112
Furst DE, Herman RA, Koehnke R, Ericksen N, Hash L, Riggs CE, Porras A, Veng-
Pedersen P (1990) Effect of aspirin and sulindac on methotrexate clearance.
J Pharm Sci 79:782-786
Gaudry SE, Sitar DS, Smyth DD, McKenzie JK, Aoki FY (1993) Gender and age as
factors in the inhibition of renal clearance of amantadine by quinine and quinidine.
Clin Pharmacol Ther 54:23-27
Gilman AG, Rall TW, Nies AS, Taylor P (eds) (1990) The pharmacological basis of
therapeutics, 8th edn. Pergamon, New York
Hager WD, Fenster P, Mayersohn M, Perrier D, Graves P, Marcus FI, Goldman S
(1979) Digoxin-quinidine interaction. N Engl J Med 300:1238-1241
Hardy BG, Zador IT, Golden L, Lalka D, Schentag JJ (1983) Effect of cimetidine on
the pharmacokinetics and pharmacodynamics of quinidine. Am J Cardiol52:172-
175
Hedaya MS, Elmquist WF, Sawchuk RJ (1990) Probenecid inhibits the metabolic
and renal clearances of zidovudine (AZT) in human volunteers. Pharm Res 7:411-
417
Hedman A, Angelin B, Arvidsson A, Dahlqvist R, Nilsson B (1990) Interactions in the
renal and biliary elimination of digoxin: stereoselective difference between qui-
nine and quinidine. Clin Pharmacol Ther 47:20-26
Hedman A, Angelin B, Arvidsson A, Beck 0, Dahlqvist R, Nilsson B, Olsson M,
Schenck-Gustafsson K (1991) Digoxin-verapamil interaction: reduction of biliary
but not renal digoxin clearance in humans. Clin Pharmacol Ther 49:256-262
Hedman A, Angelin B, Arvidsson A, Dahlqvist R (1992) Digoxin-interactions in man:
spironolactone reduces renal but not biliary digoxin clearance. Eur J Clin
Pharmacol 42:481-485
Hendel J, Nyfors A (1984) Nonlinear renal elimination kinetics of methotrexate due to
saturation of renal tubular reabsorption. Eur J Clin Pharmacol26:121-124
Higbee MD, Wood JS, Mead RA (1984) Procainamide-cimetidine interaction: a poten-
tial toxic interaction in the elderly. J Am Geriatr Soc 32:162-164
Himmelhoch JM, Poust RI, Mallinger AG, Hanin I, Neil JF (1977) Adjustment of
lithium dose during lithium-chlorothiazide therapy. Clin Pharmacol Ther 22:225-
227
Homeida M, Roberts C, Branch RA (1977) Influence of probenecid and
spironolactone on furosemide kinetics and dynamics in man. Clin Pharmacol Ther
22:402-409
HonariJ, Blair AD, Cutler RE (1977) Effects of probenecid on furosemide kinetics and
natriuresis in man. Clin Pharmacol Ther 22:395-401
Hooymans PM, Merkus FWMH (1978) Effect of quinidine on plasma concentration of
digoxin. Br Med J 2:1022
Howell SB, Olshen RA, Rice JA (1979) Effect of probenecid on cerebrospinal fluid
methotrexate kinetics. Clin Pharmacol Ther 26:641-646
Drug Interactions Involving Renal Excretory Mechanisms 209

Inotsume N, Nishimura M, Nakano M, Fujiyama S, Sato T (1990) The inhibitory effect

of probenecid on renal excretion of famotidine in young, healthy volunteers. J Clin
Pharmacol 30:50-56
Israel KS, Black HR, Nelson RL, Brunson MK, Nash JF, Brier GL, Wolney JD (1978)
Cinoxacin: pharmacokinetics and the effect of probenecid. J Clin Pharmacol
18:491-499
Jefferson JW, Kalin NH (1979) Serum lithium levels and long-term diuretic use. JAMA
241:1134-1136
Kau ST (1978) Handling of triamterene by the isolated perfused rat kidney. J
Pharmacol Exp Ther 206:701-709
Kerry RJ, Ludlow JM, Owen G (1980) Diuretics are dangerous with lithium. Br Med
J 2:371
Kerry RJ, Owen G, Michaelson S (1983) Possible toxic interaction between lithium and
piroxicam. Lancet 1:418-419
Khan IH (1991) Lithium and non-steroidal anti-inflammatory drugs. Br Med J
302:1537-1538
Klotz U, Arvela P, Rosenkranz B (1985) Famotidine, a new Hz-receptor antagonist,
does not affect hepatic elimination of diazepam or tubular secretion of
procainamide. Eur J Clin Pharmacol 28:671-675
Koren G (1987) Clinical pharmacokinetic significance of the renal tubular secretion of
digoxin. Clin Pharmacokinet 13:334-343
Kosoglou T, Rocci ML, Vlasses PH (1988) Trimethoprim alters the disposition of
procainamide and N-acetylprocainamide. Clin Pharmacol Ther 44:467-477
Kristoff CA, Hayes PE, Barr WH, Small RE, Townsend RJ, Ettigi PG (1986) Effect of
ibuprofen on lithium plasma and red blood cell concentrations. Clin Pharm 5:51-
55
Lai MY, Jiang FM, Chung CH, Chen HC, Chao PDL (1988) Dose dependent effect of
cimetidine on procainamide disposition in man. Int J Clin Pharmacol Ther 26:118-
121
Laskin OL, de Miranda P, King DH, Page DA, Longstreth JA, Rocco L, Lietman PS
(1982) Effects of probenecid on the pharmacokinetics and elimination of acyclovir
in humans. Antimicrob Agents Chemother 21:804-807
Leahey EB, Bigger JT, Butler VP, Reiffel JA, O'Connell GC, Scaffidi LE, Rottman IN
(1981) Quinidine-digoxin interaction: time course and pharmacokinetics. Am J
CardioI48:1141-1146
Leftwich AB, Walker LA, Ragheb M, Oates JA, Frolich JC (1978) Inhibition of
prostaglandin synthesis increases plasma lithium levels. Clin Res 34:291A
Levy ST, Forrest IN, Heninger GB (1973) Lithium-induced diabetes insipidus: manic
symptoms, brain and electrolyte correlates, and chlorothiazide treatment. Am J
Psychiatry 130:1014-1018
Liegler DG, Henderson ES, Hahn MA, Oliverio VT (1969) The effect of organic acids
on renal clearance of methotrexate in man. Clin Pharmacol Ther 10:849-857
Maier WP, Leon-Perez R, Miller SB (1986) Pneumonia during low-dose methotrexate
therapy. Arch Intern Med 146:602-603
May DC, Jarboe CH (1981) Inhibition of clearance of dyphylline by probenecid. N
Engl J Med 304:791
Mayall B, Poggi G, Parkin JD (1991) Neutropenia due to low-dose methotrexate
therapy for psoriasis and rheumatoid arthritis may be fatal. Med J Aust 155:480-
484
McKinney TD, Speeg KV Jr (1982) Cimetidine and procainamide secretion by proxi-
mal tubules in vitro. Am J Physiol 242:F672-F680
McSwiggan C (1978) Interaction of lithium and bicarbonate. Med J Aust 1:38
M011er JV, Sheikh MI (1982) Renal organic anion transport system: pharmacological,
physiological, and biochemical aspects. Pharmacol Rev 34:315-358
Muirhead MR, Somogyi AA, Rolan PE, Bochner F (1986) Effect of cimetidine on
renal and hepatic drug elimination: studies with triamterene. Clin Pharmacol Ther
40:400-407
210 A. SOMOGYI

Muirhead M, Bochner F, Somogyi A (1988) Pharmacokinetic drug interactions be-

tween triamterene and ranitidine in humans: alterations in renal and hepatic
clearances and gastrointestinal absorption. 1 Pharmacol Exp Ther 244:734-739
Mungall DR, Robichaux RP, Perry W, Scott lW, Robinson A, Burelle T, Hurst D
(1980) Effects of quinidine on serum digoxin concentration. Ann Intern Med
93:689-693
Nademanee K, Kannan R, Hendrickson 1, Ookhtens M, Kay I, Singh BN (1984)
Amiodarone-digoxin interaction: clinical significance, time course of develop-
ment, potential pharmacokinetic mechanisms and therapeutic implications. 1 Am
ColI CardioI4:111-116
Neuvonen PI, Karkkainen S (1983) Effects of charcoal, sodium bicarbonate and am-
monium chloride on chlorpropamide kinetics. Clin Pharmacol Ther 33:386-393
Nirenberg A, Mosende C, Mehta BM, Gisolfi AL, Rosen G (1977) High-dose
methotrexate with citrovorum factor rescue: predictive value of serum
methotrexate concentrations and corrective measures to avert toxicity. Cancer
Treat Rep 61:779-783
Overbosch D, Van Gulpen C, Hermans 1, Mattie H (1988) The effect of probenecid on
the renal tubular excretion of benzylpenicillin. Br 1 Clin PharmacoI25:51-58
Pedersen KE, Dorph-Pedersen A, Hvidt S, Klitgaard NA, Nielsen-Kudsk F (1981)
Digoxin-verapamil interaction. Clin Pharmacol Ther 30:311-316
Pedersen KE, Dorph-Pedersen A, Hvidt S, Klitgaard NA, Pedersen KK (1982) The
long-term effect of verapamil on plasma digoxin concentration and renal digoxin
clearance in healthy sUbjects. Eur 1 Clin PharmacoI22:123-127
Perel 1M, Dayton PG, Snell MM, Yti TF, Gutman AB (1969) Studies of interactions
among drugs in man at the renal level: probenecid and sulfinpyrazone. Clin
Pharmacol Ther 10:834-840
Perry PI, Calloway RA, Cook BL, Smith RE (1984) Theophylline precipitated alter-
ations of lithium clearance. Acta Psychiatr Scand 69:528-537
Peters L (1960) Renal tubular excretion of organic bases. Pharmacol Rev 12:1-35
Petersen V, Hvidt S, Thomsen K, Schou M (1974) Effect of prolonged thiazide treat-
ment on renal lithium clearance. Br Med 1 2:143-145
Pitkin D, Dubb 1, Actor P, Alexander F, Ehrlich S, Familiar R, Stote R (1981) Kinetics
and renal handling of cefonicid. Clin Pharmacol Ther 30:587-593
Ragheb MA (1987a) Ibuprofen can increase serum lithium level in lithium-treated
patients. 1 Clin Psychiatry 48:161-163
Ragheb MA (1987b) Aspirin does not significantly affect patients' serum lithium levels.
1 Clin Psychiatry 48:425
Ragheb M (1990) The interaction of lithium with phenylbutazone in bipolar affective
patients. 1 Clin PsychopharmacoI1O:149-150
Ragheb M, Powell AL (1986) Lithium interaction with sulindac and naproxen. 1 Clin
PsychopharmacoI6:150-154
Ragheb M, Ban TA, Buchanan D (1980) Interaction of indomethacin and ibuprofen
with lithium in manic patients under a steady-state lithium level. 1 Clin Psychiatry
41:397-398
Reimann IW, FrOlich lC (1981) Effects of diclofenac on lithium kinetics. Clin
Pharmacol Ther 30:348-352
Reimann IW, Diener U, FrOlich lC (1983) Indomethacin but not aspirin increases
plasma lithium ion levels. Arch Gen Psychiatry 40:283-286
Reimann IW, Golbs E, Fischer C, Frolich lC (1985) Influence of intravenous acetylsali-
cylic acid and sodium salicylate on human renal function and lithium clearance.
Eur 1 Clin PharmacoI29:435-441
Rennick BR (1972) Renal excretion of drugs: tubular transport and metabolism. Annu
Rev Pharmacol ToxicoI12:141-156
Rennick BR (1981) Renal tubule transport of organic cations. Am 1 PhysioI240:F83-
F89
Roberts DH, Kendall Ml, lack DB, Welling PG (1981) Pharmacokinetics of
cephradine given intravenously with and without probenecid. Br 1 Clin Pharmacol
11:561-564
Drug Interactions Involving Renal Excretory Mechanisms 211

Rodin SM, Johnson BF (1988) Pharmacokinetic interactions with digoxin. Clin

Pharmacokinet 15:227-244
Rodriguez N, Madsen PO, Welling PG (1979) Influence of probenecid on serum levels
and urinary excretion of cinoxacin. Antimicrob Agents Chemother 15:465-469
Rodvold KA, Paloucek FP, Jung D, Gallastegui J (1987) Interaction of steady-state
procainamide with H 2-receptor antagonists cimetidine and ranitidine. Ther Drug
Monit 9:378-383
Sand TE, Jacobsen S (1981) Effect of urine pH and flow on renal clearance of
methotrexate. Eur J Clin Pharmacol 19:453-456
Schenck-Gustafsson K, Dahlqvist R (1981) Pharmacokinetics of digoxin in patients
subjected to the quinidine-digoxin interaction. Br J Clin Pharmacol11:181-186
Sierles FS, Ossowski MG (1982) Concurrent use of theophylline and lithium in a
patient with chronic obstructive lung disease and bipolar disorder. Am J Psychia-
try 139:117-118
Singh RR, Malaviya AN, Pandey IN, Guleria JS (1986) Fatal interaction between
methotrexate and naproxen. Lancet 1:1390
Somogyi A, Bochner F (1984) Dose and concentration dependent effect of ranitidine
on procainamide disposition and renal clearance in man. Br J Clin Pharmacol
18:175-181
Somogyi A, McLean A, Heinzow B (1983) Cimetidine-procainamide pharmacokinetic
interaction in man: evidence of competition for tubular secretion of basic drugs.
Eur J Clin Pharmacol 25:339-345
Somogyi A, Stockley C, Keal J, Rolan P, Bochner F (1987) Reduction of metformin
renal tubular secretion by cimetidine in man. Br J Clin Pharmacol 23:545-551
Somogyi AA, Hovens CM, Muirhead MR, Bochner F (1989) Renal tubular secretion
of amiloride and its inhibition by cimetidine in humans and in an animal model.
Drug Metab Dispos 17:190-196
Somogyi AA, Bochner F, Sallustio BC (1992) Stereoselective inhibition of pindolol
renal clearance by cimetidine in humans. Clin Pharmacol Ther 51:379-387
St Clair EW, Rice JR, Snyderman R (1985) Pneumonitis complicating low-dose
methotrexate therapy in rheumatoid arthritis. Arch Intern Med 145:2035-2038
Steiness E (1974) Renal tubular secretion of digoxin. Circulation 50:103-107
Stewart CF, Fleming RA, Magneson P, Germain B, Evans WE (1990a) Effect of aspirin
(AS A) on the disposition of methotrexate (MTX) in patients with rheumatoid
arthritis (RA). Clin Pharmacol Ther 47:139
Stewart CF, Fleming RA, Arkin CR, Evans WE (1990b) Coadministration of naproxen
and low-dose methotrexate in patients with rheumatoid arthritis. Clin Pharmacol
Ther 47:540-546
Stockley IH (1994) Drug interactions. A source book of adverse interactions, their
mechanisms, clinical importance and management, 3rd edn. Blackwell, Oxford
Thomsen K, Schou M (1968) Renal lithium excretion in man. Am J Physiol 215:823-
827
Thyss A, Milano G, Kubar J, Namer M, Schneider M (1986) Clinical and pharmacoki-
netic evidence of a life-threatening interaction between methotrexate and
ketoprofen. Lancet 1:256-258
Tracy TS, Krohn K, Jones DR, Bradley JD, Hall SD, Brater DC (1992) The effects of
a salicylate, ibuprofen, and naproxen on the disposition of methotrexate in pa-
tients with rheumatoid arthritis. Eur J Clin PharmacoI42:121-125
Ullrich KJ (1994) Specificity of transporters for "organic anions" and "organic cations"
in the kidney. Biochim Biophys Acta 1197:45-62
Ullrich KJ, Fritzsch G, Rumrich G, David C (1994) Polysubstrates: substances that
interact with renal contraluminal P AH, sulfate, and NMeN transport: sulfamoyl-
sulfonylurea-, thiazide- and benzene amino-carboxylate (nicotinate) compounds.
J Pharmacol Exp Ther 269:684-692
Van Crugten J, Bochner F, Keal J, Somogyi A (1986) Selectivity of the cimetidine-
induced alterations in the renal handling of organic substrates in humans.
Studies with anionic, cationic and zwitterionic drugs. J Pharmacol Exp Ther
236:481-487
212 A. SOMOGYI: Drug Interactions Involving Renal Excretory Mechanisms

Veenendaal RJ, Brooks PM, Meffin PJ (1981) Probenecid-clofibrate interaction. Clin

Pharmacol Ther 29:351-358
Vlasses PH, Kosoglou T, Chase SL, Greenspon AJ, Lottes S, Andress E, Ferguson RK,
Rocci ML (1989) Trimethoprim inhibition of the renal clearance of procainamide
and N-acetylprocainamide. Arch Intern Med 149:1350-1353
Waldorff S, Andersen JD, Heebl/lll-Nielsen N, Nielsen OG, Moltke E, SI/lrensen U,
Steiness E (1978) Spironolactone-induced changes in digoxin kinetics. Clin
Pharmacol Ther 24:162-167
Waldorff S, Hansen PB, Kjaergard H, Buch J, Egeblad H, Steiness E (1981) Amiloride-
induced changes in digoxin dynamics and kinetics: abolition of digoxin-induced
inotropism with amiloride. Clin Pharmacol Ther 30:172-176
Waldorff S, Hansen PB, Egeblad H, Berning J, Buch J, Kjaergard H, Steiness E (1983)
Interactions between digoxin and potassium-sparing diuretics. Clin Pharmacol
Ther 33:418-423
Waller ES, Sharanevych MA, Yakatan GJ (1982) The effect of probenecid on nafcillin
disposition. J Clin Pharmacol 22:482-489
Weidekamm E, Portmann R, Suter K, Partos C, Dell D, Lucker PW (1987) Single- and
multiple-dose pharmacokinetics of fteroxacin, a triftuorinated quinolone, in hu-
mans. Antimicrob Agents Chemother 31:1909-1914
Weiner 1M, Mudge GH (1964) Renal tubular mechanisms for excretion of organic
acids and bases. Am J Med 36:743-762
Windle J, Prystowsky EN, Miles WM, Heger JJ (1987) Pharmacokinetic and electro-
physiologic interactions of amiodarone and procainamide. Clin Pharmacol Ther
41:603-610
Wingender W, Beermann D, Foster D (1985) Mechanism of renal excretion of
ciproftoxacin (Bay 9867), a new quinolone carboxylic derivative, in humans. Che-
motherapy 4 [Suppl 2]:403-404
Section II
Pharmacodynamic Drug Interactions
CHAPTER 7
Drug-Drug Interactions at Receptors
and Other Active Sites
M. SCHORDERET and J.D. FERRERO

A. Introduction
Pharmacodynamic interactions occur at receptor sites or result more broadly
from the effect of the drugs on physiological processes. Both antagonism and
synergism may be produced. The synergistic interactions, as far as they have
the potential to be beneficial and clinically useful, are reviewed separately in
this handbook (see Chap. 8).
The present chapter will concentrate on adverse drug interactions result-
ing in either a diminished therapeutic effect or an increased toxicity. Its aim is
to describe the mechanisms by which pharmacodynamic interactions may
occur and to illustrate them with clinically relevant examples. It is not a
compilation of all possible and known interactions, duplicating standard
publications (SCHOU 1982; MciNNES and BRODIE 1988; STOCKLEY 1991;
BUGNON et al. 1993; HANSTEN et al. 1993). Accordingly, the classification
adopted here is based on sites of action and not on drug classes. Different steps
in the same process, e.g., synaptic transmission, or different mechanisms tak-
ing part in the regulation of one system or function, e.g., blood pressure, may
be affected.
Since they are based on mechanisms of action supposedly known to
the prescriptor, pharmacodynamic interactions should be more easily pre-
vented and should have less clinical impact than pharmacokinetic interactions.
However, the relevance of such a statement may be questioned in view of
the recent advances in the characterization and detection of kinetic
interactions.

B. Mechanisms of Pharmacodynamic Interactions

I. Transmitter Systems

1. Noradrenergic Synapse
The various drugs interfering with practically every step of the nor adrenergic
transmission represent a well-established source of pharmacodynamic interac-
tions (Fig. 1). In the CNS, where adrenaline is a neurotransmitter on its own
216 M. SCHORDERET and J.D. FERRERO

(MEFFORD 1988; COOPER et al. 1991), similar interactions can occur. The spe-
cific plasma membrane noradrenaline transporter has been cloned and charac-
terized (PACHOLCZYK et al. 1991; USDIN et al. 1991). It plays a role in the
mechanism of action and potential interactions of amphetamine and tyramine,
which stimulate the outward transport of noradrenaline, as well as of cocaine
and imipramine, which block the inward transport (reuptake) of noradrena-
line (Fig. 1).
Possible interactions at postsynaptic receptors are not shown in Fig. 1,
since they can easily be anticipated. Patients on ,B-blockers are particularly at
risk (Table 1). For example, the blockade of ,B-receptors makes a patient
resistant to adrenaline in the emergency treatment of anaphylaxis. In this

TYROSINE L-DOPA

+
DOPAMINE

+
NORADRENALINE

i:~::\
~ ~
~G\!Y

/RES,

--e--- MAO[-A

• • • COC AMPH/TYR
• • ~

Fig. 1. Noradrenergic synapse. Pharmacodynamic interactions at presynaptic sites

and transporters (T). AMPH, amphetamine; COC, cocaine; CLO, clonidine; GVA,
guanethidine; IMI, imipramine and imipramine-like antidepressants; MAOI-A,
monoamine oxidase inhibitor (type A); MD, a-methyldopa; RES, reserpine; TYR,
tyramine. The localization of drugs (inside or outside the presynaptic terminal) is
purely arbitrary. For the sake of clarity, all subtypes of receptors (a and {J) are gathered
together on the same postsynaptic membrane. Symbols EEl and 8 on the bidirectional
arrows represent mutual potentiation and inhibition, respectively, of the effects of the
compounds at both ends of the arrows (see text for additional explanations)
Drug-Drug Interactions at Receptors and Other Active Sites 217

situation, glucagon would still be effective as an inotropic agent (BOCHNER and

LICHTENSTEIN 1991). Similarly, the early symptoms of hypoglycaemia may be
masked in a diabetic patient treated with insulin or oral hypoglycaemic agents
(Table 1; see also Sect. B.III.2). Unopposed a-adrenergic stimulation is the
probable mechanism of the increased cardiovascular toxicity (hypertension,
coronary constriction, cf. Table 1) of cocaine (LANGE et al. 1990).
As expected, drugs facilitating the noradrenergic transmission [amphet-
amine, cocaine, monoamine oxidase inhibitors (MAO Is ) and imipramine-like
antidepressants] will potentiate the effects of lX:i- or f3112-adrenergic receptor
agonists. Indeed, the dosage of adrenaline in the treatment of anaphylaxis

Table 1. Pharmacodynamic interactions with f3-blockers

Associated drugs Mechanisms Clinical consequences

llt-Agonists llt -Receptor-induced Hypertensive crises,

and indirect vasoconstriction unopposed coronary vasoconstriction
sympathomimetics by fJz-receptor-induced
(adrenaline, vasodilation
phenylephrine,
etilefrine, dopamine,
cocaine)
Clonidine Increased sympathetic activity: Increased rebound
withdrawal llt -receptor-induced hypertension
vasoconstriction unopposed
by fJz-receptor-induced
vasodilation
Ergotamine llt- and 5-HT2-receptor-induced Peripheral vasospasm
vasoconstriction, unopposed by
fJz-receptor-induced vasodilation
llt-Antagonists Potentiation of the "first-dose Orthostatic hypotension
(prazosin) effect" of prazosin
Digoxin Additive slowing of conduction Bradycardia
Disopyramide Additive cardiodepressant Bradycardia, heart block,
activity hypotension
Verapamil, Additive cardiodepressant Bradycardia, heart block,
diltiazem activity hypotension, left
ventricular failure
NSAID Inhibition of cyclooxygenases Decreased
antihypertensive
effect of J3-blockers
Insulina Hypoglycaemia; masking of Glycaemia control in
hypoglycaemia symptoms diabetic patients made
more difficult

NSAID, non-steroidal anti-inflammatory drugs; see Sect. B.IV.2.

aSee Sect. B.1II.2.
218 M. SCHORDERET and J.D. FERRERO

should be strongly reduced in a patient receiving tricyclic antidepressants

(STOCKLEY 1991). Similarly, sympathomimetic OTC preparations containing,
for example, phenylpropanolamine or imidazolines as systemic or topical
nasal decongestants can cause potentially serious reactions (CETARUK and
AARON 1994). In contrast, the indirect sympathomimetic agents mentioned
above will attenuate the effects of sympatholytic drugs such as clonidine
and a-methlydopa (agonists at central az-adrenergic receptors that inhibit
the sympathetic tone), reserpine (a blocker of the vesicular catecholamine
transporter) and guanethidine. Since the latter agent must be taken up in the
nerve terminal through the membrane noradrenaline transporter in order to
be active, it will lose its antihypertensive effect in the presence of imipramine
(Fig. 1) or other drugs that block the reuptake of noradrenaline (MICHELL et
al. 1970). The inhibition by imipramine of the antihypertensive effect of
clonidine or a-methyldopa is based on a different mechanism, possibly a direct
interaction at central az-receptors.
It has been known for several decades that the combination of MAOIs
with foods and beverages containing tyramine may induce the "cheese reac-
tion", characterized by hypertension, headache, sweating, chest pain and even
cerebral haemorrhage and death (BLACKWELL and MARLEY 1966). This reac-
tion results from tyramine getting free access, in the absence of MAO activity,
to the adrenergic nerve terminals and varicosities, where it stimulates norad-
renaline release. It appears, however, that the risk of serious reactions with
dietary tyramine is much reduced with new selective and reversible MAOIs-A
such as moclobemide (BERLIN et al. 1989; FITTON et al. 1992). Since tyramine is
a substrate for MAO-A, the reversibility of MAO inhibition may be a more
determinant property than the selectivity (GIESCHKE et al. 1988). Moreover,
the simultaneous administration of older irreversible MAOIs with tricyclic
antidepressants has been shown to produce a potentially lethal central
and cardiovascular syndrome, which is likely to be due to increased amine
concentrations at postsynaptic receptors (SPINA and PERUCCA 1994). The
association of moclobemide with tricyclic antidepressants, if the treatment is
initiated with slowly increasing doses of both drugs, or if imipramine is
started before moclobemide, is probably safe, but remains to be fully charac-
terized (FITTON et al. 1992; SPINA and PERUCCA 1994). Finally, possible
pharmacodynamic interactions between MAOIs and pethidine (meperidine)
or dextromethorphane are described in Sect. B.I.3.
In considering the numerous possible interactions of tricyclic antidepres-
sants, it should be remembered that besides blocking amine membrane
transporters, they possess additional (and to some extent paradoxical)
antagonistic properties at lXt-adrenergic, HI-histamine and muscarinic
receptors (SUGRUE 1981; REMICK 1988), and that they exert quinidine-like
effects on the membrane of excitable cells (GLASSMAN and BIGGER 1981).
In this respect, the "second-generation" antidepressants, particularly the
serotonin-selective reuptake inhibitors (SSRIs, see Sect. B.I.3), appear to be
much safer.
Drug-Drug Interactions at Receptors and Other Active Sites 219

2. Dopaminergic Synapse
Dopamine is a neurotransmitter in the eNS, and as a circulating catechola-
mine released from the adrenals it also plays multiple roles in the periphery
(HORN and MURPHY 1991). In addition, brain dopamine receptors are the
primary target in the treatment of schizophrenia, Parkinson's disease and
Huntington's chorea (SEEMAN and VAN TOL 1994). A better knowledge of the
processes involved in the dopaminergic transmission, as well as of the pharma-
cological characteristics of the subtypes of receptors (D! to D s), may help to
understand and possibly to avoid drug-drug interactions (Fig. 2).
The interaction between L-dopa and pharmacological doses of pyridoxine
(vitamin B 6 ) is due to an increased peripheral decarboxylation of L-dopa by

TYROSINE --........,__ L-DOPA

PYRIDOXINE

+
(VITAMIN 86)

DOPAMINE

t
RES

MAOI- B

COC AMPH/TYR

Fig. 2. Dopaminergic synapse. Pharmacodynamic interactions at presynaptic sites and

transporter (T). AMPH, amphetamine; CDC, cocaine; MAOI-B, monoamine oxidase
inhibitor (type B); R, reserpine; TYR, tyramine. The localization of drugs (inside or
outside the presynaptic terminal) is purely arbitrary. For the sake of clarity, all subtypes
of receptors (D!-Ds) are gathered together on the same postsynaptic membrane.
CLOZ, clozapine; NLT, typical neuroleptics; SULP, sulpiride. Symbols Ell and 8 on
the bidirectional arrows represent mutual potentiation and inhibition, respectively, of
the effects of the compounds at both ends of the arrows (see text for additional
explanations)
220 M. SCHORDERET and J.D. FERRERO

the L-amino acid decarboxylase cofactor, which results in a decreased transfer

of L-dopa to the eNS and a concomitant reduction in therapeutic efficacy. The
interaction does not occur when L-dopa is co-administered with a peripheral
inhibitor of the L-amino acid decarboxylase (MARS 1974). The specific plasma
membrane transporter of dopamine has been cloned and characterized (GIROS
et al. 1991; KILTY et al. 1991; SHIMADA et al. 1991). It is concerned with the
mechanism of action and potential interactions of amphetamine and tyramine,
which stimulate the outward transport of dopamine, as well as of cocaine,
which blocks the inward transport (reuptake) of dopamine (Fig. 2). Possible
interactions at postsynaptic receptors are not shown in Fig. 2, since they
can easily be anticipated. As expected, the typical neuroleptics represented
by chlorpromazine, flupenthixol or haloperidol, which are potent antagonists
at Dl and D2 receptor subtypes, will inhibit the therapeutic action of the
dopamine precursor (L-dopa) and of amantadine, as well as of preferential D z
agonists (e.g., bromocriptine and analogs) in the treatment of Parkinson's
disease. Further interactions with L-dopa can also be expected from the mixed
D z/D 3 antagonist sulpiride. However, D z antagonists that exhibit a low inci-
dence of extrapyramidal effects, such as domperidone or metoclopramide,
may be of benefit in the treatment of L-dopa- or Dz-agonist-induced emesis
(MITCHELSON 1992). Though theoretically founded, interactions implying the
D 3- (e.g., selective agonists or antagonists being able to decrease or increase
the dopaminergic transmission, respectively) or D 4-receptor subtypes (e.g., the
relatively selective D4 antagonist clozapine) do not yet seem to be clinically
relevant.
It should be noted that most neuroleptics, particularly those of the phe-
nothiazine class, may cause additional interactions due, for example, to an
antagonism at a-adrenergic and muscarinic receptors (DILSAVER 1993).

3. Serotonergic Synapse
The last decade has been marked by a rapid progress in serotonin (5-hydroxy-
tryptamine, 5-HT) research with the cloning, expression and characterization
of numerous subtypes of 5-HT receptors (BOESS and MARTIN 1994), as well as
of the specific plasma membrane transporter of serotonin (BLAKELY et al. 1991;
HOFFMAN et al. 1991).
The serotonin transporter is concerned with the mechanism of action
and possible interactions of amphetamine and 3,4-methylenedioxy-
methamphetamine (MDMA, "ecstasy"), which stimulate the outward trans-
port of serotonin, as well as of cocaine and the selective serotonin reuptake
inhibitors (SSRls), which block the inward transport of serotonin (Fig. 3). The
combined treatment with irreversible MAOls and SSRls has been shown to
cause a serotonin syndrome (FEIGHNER et al. 1990; GRAM 1994; SPINA and
PERUCCA 1994). Moreover, a lethal serotonin syndrome has been reported
after the concomitant administration of high-dose moclobemide with
citalopram or clomipramine (NEUVONEN et al. 1993; MITCHELL 1994). Thus, the
Drug-Drug Interactions at Receptors and Other Active Sites 221

TRYPTOPHAN _ _ _ +- 5-HYDROXY -
TRYPTOPHAN

+
5-HYDROXYTRYPT AMINE
(5 - HT)
(SEROTONIN)

SSRI ---<D---- MAOI-A

•• COC AMPH/MDMA
~

Fig. 3. Serotonergic synapse. Pharmacodynamic interactions at presynaptic sites

and transporters (1). AMPH, amphetamine; CDC, cocaine; MDMA, 3,4-meth-
ylenedioxymethamphetamine ("ecstasy"); MAOI-A, monomaine oxidase inhibitor
(type A); RES, reserpine; SSRI, selective serotonin reuptake inhibitors, such as
citalopram, fiuoxetine, fiuvoxamine, paroxetine, sertraline and zimeldine. The localiza-
tion of drugs (inside or outside the presynaptic terminal) is purely arbitrary. For the
sake of clarity, different subtypes of receptors (5-HT, to 5-HT4) are gathered together
on the same postsynaptic membrane. The unlabelled postsynaptic receptor signals that
additional subtypes (5-HTs to 5-HT7 ), for which ligands remain to be found, have been
characterized. Symbols Ef:) and e on the bidirectional arrows represent mutual potentia-
tion and inhibition, respectively, of the effects of the compounds at both ends of the
arrows (see text for additional explanations)

prospective benefits of combining SSRIs with MAOIs - even with the revers-
ible MAO I-A moclobemide - in the treatment of depression should be seri-
ously weighed against the associated risks (SPINA and PERUCCA 1994).
Like cocaine, the OTC antitussive dextromethorphan, as well as pethidine
(meperidine), can also block the neuronal uptake of serotonin (BROWNE and
LINTER 1987; STACK et al. 1988; HILL et al. 1992; CETARUK and AARON 1994).
The association of dextromethorphan or pethidine with MAOIs (including
222 M. SCHORDERET and J.D. FERRERO

moclobemide) has been shown to produce clinical symptoms very similar to

those of the serotonin syndrome (LEJOYEUX et al. 1994), with hypertensive
crises, hyperpyrexia, hyperexcitation, muscle rigidity, coma and death
(AMREIN et al. 1992; CETARUK and AARON 1994; SPINA and PERUCCA 1994). A
serotonin syndrome has also been reported when the 5-HT precursor tryp-
tophan was associated with MAOIs, SSRIs or imipramine-like drugs
(LEJOYEUX et al. 1994), as well as during lithium therapy associated with
fluvoxamine or fluoxetine (GUY 1992; SPINA and PERUCCA 1994).
Possible pharmacodynamic interactions at the postsynaptic level would
result from the erroneous simultaneous administration of agonists and antago-
nists acting on the same 5-HT receptor subtype (Fig. 3) or from the association
of poorly selective or non-selective agonists and antagonists. In view of the
increased risk of peripheral vasospasm, for example, ergotamine and
sumatriptan should not be administered concomitantly in the therapy of
migraine, nor should they be prescribed together with MAOIs or SSRIs
(DRUG and THERAPEUTICS BULLETIN 1992).

4. Cholinergic Synapse
In the cholinergic transmission, unlike the monoaminergic transmissions,
drugs affecting the synthesis (e.g., choline, lecithins), storage and uptake of the

ACET)
CHOLINE
~~r--- UPTAKE

RELEASE
INHIBITORS

NON DEPOLARIZING • • •
NEUROMUSCULAR • •
BLOCKING DRUGS

CHOLINESTERASE
INHIBITORS

Fig.4. Cholinergic synapse. Pharmacodynamic interactions of non-depolarizing neuro-

muscular-blocking drugs with cholinesterase inhibitors and inhibitors of acetylcholine
release (see text for details). Symbols EB and e on the bidirectional arrows represent
mutual potentiation and inhibition, respectively, of the effects of the compounds at
both ends of the arrows. M, muscarinic receptors; N, nicotinic receptors; AChE, acetyl-
cholinesterase
Drug-Drug Interactions at Receptors and Other Active Sites 223

Table 2. Drugs with anticholinergic (antimuscarinic) activity

Therapeutic classes Examples

Antiarrhythmic drugs Quinidine

Antiasthmatics Ipratropium
Antidepressants Imipramine
Antiemetics (motion sickness) Scopolamine
Antiepileptics Carbamazepine
Antiparkinsonians Trihexyphenidyl
Antiulcer drugs Pirenzepine
Hcantihistaminics Promethazine
Mydriatics Homatropine
Neuroleptics Chlorpromazine
Parasympatholytics Atropine

transmitter have not been found therapeutically useful. A few drugs, however,
have been shown to interfere presynaptically with acetylcholine release at
the neuromuscular junction (Fig. 4). Apart from the botulinum toxins, some
antibiotics (aminoglycosides, polymyxins, lincosamides) (SINGH et al. 1982),
quinidine, quinine and procainamide, as well as magnesium salts, playa role in
this effect (CLARKE adn MIRAKHUR 1994). The aminoglycosides are not only
able to produce weakness on their own or aggravate the muscular symptoms in
myasthenia gravis (DRACHMAN 1994), they may also enhance the effect of the
neuromuscular blocking agents (CLARKE and MIRAKHUR 1994).
Most interactions occurring at cholinergic receptors involve additive
effects between muscarinic receptor blocking drugs, examples of which can be
found in many therapeutic classes (Table 2). In the nicotinic domain, additive
effects with dangerous cardiovascular consequences have also been reported
in patients treated with transdermal nicotine delivery systems who do not stop
smoking or who start smoking again (FIORE et al. 1992).
Cholinesterase inhibitors are routinely used in anaesthesiology to reverse
curare-induced neuromuscular block (Fig. 4). As the use of cholinesterase
inhibitors, notably tacrine, has been proposed in the treatment of Alzheimer's
disease, it should be remembered that any agent with antimuscarinic proper-
ties which crosses the blood-brain barrier will interact negatively. A more
insidious interaction of anticholinergic drugs is the inhibition of the
gastrokinetic effect of the 5-HT4 agonists metoclopramide and cisapride that
act by stimulating acetylcholine release in the enteric nervous system
(SCHOURKES et al. 1988).

5. GABAergic Synapse
GAB A is the main inhibitory transmitter in the mammalian brain. In postsyn-
aptic neurons, it produces either a fast response, through the ligand-gated
ion-channel GABAAreceptor, or a slow modulatory response, through the G-
protein-coupled GABAB receptor (MODY et al. 1994). As shown in Table 3,
various substances can either facilitate or inhibit the GABAA-mediated
increase in CI- conductance (SIEGHART 1992).
224 M. SCHORDERET and J.D. FERRERO

Table 3. Modulation of GABAergic neurotransmission

coupled to GABA A receptors

Drugs which reinforce Drugs which inhibit

GABA-mediated processes GAB A-mediated processes

Benzodiazepines Flumazenila
Barbiturates
Gabapentin Penicillins
Progabide Quinolones
Vigabatrin
Ethanol Bicuculline
General anaesthetics Picrotoxin
A vermectin insecticides Cyclodiene insecticides

a Benzodiazepine receptor antagonist.

Benzodiazepines, barbiturates and probably ethanol are allosteric modu-

lators that reinforce GABAergic transmission. Potentiation of CNS-depres-
sant effects, with an increased risk of respiratory depression, is a well-known
example of pharmacodynamic interaction brought about by the combination
of any of these hypnosedative agents. Inversely, the administration of
flumazenil, and antagonist at the benzodiazepine site of the GABA A receptor,
will precipitate an acute withdrawal syndrome in a patient physically depen-
dent on benzodiazepines. Convulsions may be triggered, particularly in the
epileptic patient treated with benzodiazepines (MARCHANT et al. 1989).
Among other substances of experimental or toxicological interest (Table
3), penicillins in high doses and quinolone antibiotics have been reported to
cause CNS stimulation or convulsions, which may be due to the inhibition
of GABAergic processes (CURTIS et al. 1972; HALLIWELL et al. 1993). The
non-steroidal anti-inflammatory drugs, fenbufen in particular, may increase
the incidence of quinolone-induced convulsions. The mechanism involved in
this interaction remains unknown (CHRIST et al. 1988).
As a GABAB receptor agonist, baclofen may increase or decrease the
level of synaptic inhibition in the CNS, depending on whether the receptors
are located pre- or postsynaptically (MOTT and LEWIS 1994). No clinically
important pharmacodynamic interaction has yet been reported with
baclofen.

II. Ion Channels

1. Cardiac Ion Channels and Antiarrhythmic Drugs

It is well established that class I antiarrhythmic drugs have sodium channel
blocking properties. The affinity for or access to the channel receptor site may
depend on specific states of the channel. Most drugs have low affinity for the
Drug-Drug Interactions at Receptors and Other Active Sites 225

sodium channel in the resting state and the affinity increases when the channel
is in the activated (quinidine) or inactivated (lidocaine) states. Class I agents
are further characterized by the association/dissociation kinetics to and from
the channel receptor site (WOOSLEY 1991).
The concomitant administration of antiarrhythmic drugs with distinct
properties (e.g., class Ia/lb) may result in either adverse or beneficial interac-
tions (LANGIERI et al. 1995). As an example of the latter, mexiletine (Ib) and
quinidine (Ia) in combination have been found to be more effective at a lower
dose and to produce fewer adverse effects than each drug given alone in higher
dosage (BIGGER 1984; GIARDIA and WESCHSLER 1990). This may be explained
by the fact that quinidine-induced prolongation of repolarization prolongs
the inactivated state of the sodium channel. As a result, the binding of
mexiletine, which possesses higher affinity for the channel in the inactivated
state, is increased. Moreover, the competition between "fast" and "slow"
sodium channel blockers (including non-antiarrhythmic drugs) may be
effective in reversing toxic electrophysiologic effects. Lidocaine has been
shown to reverse the QRS prolongation due to cocaine (BAUMAN et al. 1994)
or propoxyphene (WHITCOMB et al. 1989) overdoses. Finally, the drugs that
cause repolarization abnormalities (see Sect. B.II.2) tend to have greater
toxicity at slower heart rate. Increasing heart rate by temporary pacing or
isoprenaline infusion is the recommended treatment in this case (BEN-DAVID
and ZIPES 1993).
Cardiac calcium channels, particularly in the atrioventricular node, are
concerned with a well-known additive interaction between class II (f3-
blockers) and class IV (verapamil, diltiazem) antiarrhythmic drugs (see also
Table 1). The resulting inhibition of atrioventricular conduction is due to the
blockade of nodal calcium channels induced both directly by the calcium
channel blocker and indirectly, through a reduction in sympathetic drive by
the f3-blocker (WINNIFORD et al. 1982).

2. Potassium Channels and Drug-Induced Torsades de Pointes

Torsades de pointes is a potentially lethal ventricular arrhythmia, which can be
produced by class Ia and class III antiarrhythmic agents, as well as by practi-
cally any drug that prolongs repolarization (BEN-DAVID and ZIPES 1993). A
delay in the opening of the repolarizing outward-rectifying potassium channels
increases the duration of the action potential and of the QT interval on
the electrocardiogram. Numerous drugs have been associated with torsades
de pointes and the long QT syndrome (Table 4). Synergism is likely to
occur when QT-modifying drugs are administered concomitantly. This is, for
example, the reason why meftoquine and halofantrine should not be used
together (WHITE 1994). Electrolyte disorders, such as the hypokalaemia/
hypomagnesaemia caused by the administration of diuretics, can precipitate
torsades de pointes and have a synergistic effect with class Ia antiarrhythmic
drugs (SISCOVICK et al. 1994). It should be noted that apparently "safe" drugs,
226 M. SCHORDERET and J.D. FERRERO

Table 4. Some drugs associated with torsades de pointes

Therapeutic classes Examples

Antiarrhythmic drugs
- Class Ia Quinidine
- Class III Sotalol
Antihypertensive agents Ketanserine
Antimalarial drugs Chloroquine, mefioquine, halofantrine
CNS stimulants 3,4-Methylenedioxy-methamphetamine ("ecstasy")
Diuretics Hydrochlorothiazide
Hj-antihistaminics Terfenadine, astemizole
Macrolide antibiotics Erythromycin
Neuroleptics Chlorpromazine, haloperidol
Tricyclic antidepressants Imipramine

like terfenadine and astemizole, can create a life-threatening risk when admin-
istered in overdose or in association with macrolide antibiotics, particularly
erythromycin. In the latter case, the interaction is largely of a pharmacokinetic
nature (see Chap. 6, this volume).

3. ATP-Sensitive Potassium Channels

ATP-sensitive K+ channels (KATP channels) of the pancreatic B cell are the
target of drugs producing opposite effects on insulin secretion. KATP channel
blockers, such as the sulfonylureas tolbutamide and glibenclamide, reduce
membrane polarization. This will increase the opening probability of
voltage-dependent Ca2+ channels resulting in insulin release (SCHMID-
ANTOMARCHI et al. 1987). On the contrary, diazoxide and somatostatin are
KATP channel openers, which hyperpolarize the cell and inhibit insulin release
(EDWARDS and WESTON 1993). The other KATP channel openers, such as
minoxidil and pinacidil, have a much lower affinity for the B cell than for
smooth muscle channels and are thus much less likely to interact with
antidiabetic sulfonylureas.

III. Hormonal Systems

Natural, semisynthetic and synthetic hormones are used as replacement
therapy for hormone deficiency states or as drugs to modulate the hypotha-
lamic-hypophyseal regulations of the gonads and other endocrine glands.
Most clinically important interactions involving hormonal treatments are of
a pharmacokinetic nature. However, a few pharmacodynamic interactions
concern the corticosteroids and the antidiabetic agents.

1. Adrenal Corticosteroids
Interactions implying the corticosteroids can be deduced from their broad
spectrum of activity. Their metabolic effects, which tend to produce a
Drug-Drug Interactions at Receptors and Other Active Sites 227

hyperglycaemic state, will reduce the efficacy of insulin and other antidiabetic
agents (see below). Those glucocorticoids possessing mineralocorticoid activ-
ity will aggravate the hypokalaemia induced by thiazide or loop diuretics.
Finally, the corticosteroids may interfere with the development of the protec-
tive immune response to vaccines (BUTLER and ROSSEN 1977; ZIELENSKI et al.
1986). Moreover, in view of altered defense mechanisms, live vaccines should
not be administered to patients treated with systemic corticosteroids (GROSS et
al. 1985).

2. Glycaemic Regulation
The association of insulin and oral hypoglycaemic agents has been advocated
in the treatment of type II diabetes (GERICH 1989; LEWITT et al. 1989). Clearly
enough, the additive effects of such a combination would increase the risk of
severe hypoglycaemia. Similarly, the a-glucosidase inhibitors (e.g., acarbose)
may aggravate insulin- and sulphonylurea-induced hypoglycaemia (BALFOUR
and McTAVISH 1993). Inversely, the therapeutic efficacy of oral antidiabetic
agents would be reduced by the simultaneous administration of thiazide
diuretics or of diazoxide (see Sect. B.I1.3).
Many drugs affect glycaemic regulation and can thus interfere with
antidiabetic agents. It has been mentioned previously (see Sect. B.Ll) that
f3-blockers may mask the early signs of hypoglycaemia and should be avoided
in diabetic patients. In addition, non-selective f3-blockers have been reported
to cause severe hypoglycaemia in both diabetic and non-diabetic patients
(KOTLER et al. 1966; ANGELO-NIELSON 1980). Unexpectedly, the salicylates in

Table 5. Drugs affecting glycaemic regulationa. Possible

interactions with insulin and oral hypoglycaemic agents

Hyperglycaemia Hypoglycaemia

f3-Blockers f3-Blockers
A-Adrenergic agonists
Cyclosporine Disopyramide
Corticosteroids Angiotensin-converting enzyme
inhibitors (ACEIs)
Diazoxide Ethanol
Nicotinic acid Fibric acid derivatives
Octreotide
Pentamidine Pentamidine
Phenothiazines Quinine
Phenytoin Salicylates
Thiazide diuretics Streptozotocin
Thyroid hormones

aSee the review by PANDIT et al. (1993) for an exhaustive

list and evaluation of medications influencing plasma
glucose levels.
228 M. SCHORDERET and J.D. FERRERO

high doses (4-6g1day) have been shown to decrease blood glucose concentra-
tions (SELTZER 1989). Salicylates increase insulin secretion in non-insulin-
dependent diabetes, increase insulin sensitivity and inhibit lipolysis (MICOSSI
et al. 1978; PANDIT et al. 1993). They can thus enhance the hypoglycaemic
effects of insulin and oral antidiabetic agents (RICHARDSON et al. 1986). Addi-
tional pharmacokinetic interactions with the sulphonylureas (e.g., displace-
ment from protein-binding sites) may also playa role (see Chap. 4, this
volume). If anti-inflammatory doses of aspirin are to be prescribed to a dia-
betic patient, they should be started low and increased progressively with
frequent blood glucose monitoring.
As mentioned in Sect. B.II!.l, the glucocorticoids represent another
cause of interaction with antidiabetic agents. By stimulating hepatic gluconeo-
genesis and decreasing peripheral utilization of glucose (JACKSON and
BRESSLER 1981; GERICH 1989), they tend to produce a diabetes-like state
(PANDIT et al. 1993).
This survey is limited to a few classical examples of pharmacodynamic
interactions that implicate insulin and other antidiabetic agents. Many
additional drugs (including ethanol) can interfere with glycaemic regulation
(Table 5).

IV. Homeostatic Regulations

1. Renal Haemodynamics and Drug-Induced Acute Renal Failure

Decreased renal perfusion is a major factor that predisposes to nephrotoxicity.
Besides various causes unrelated to drugs, renal perfusion can be decreased
by drug-induced sodium and volume depletion (diuretics) or systemic
hypotension (antihypertensive agents). In the face of reduced renal perfusion,
the glomerular hydrostatic pressure and filtration rate are maintained through
a balance between angiotensin II-mediated efferent arteriole constriction and
prostaglandin-effected afferent arteriole dilation (Fig. 5). In such a condition,
drugs interfering with either prostaglandin synthesis (e.g., the non-steroidal
anti-inflammatory drugs, NSAIDs) or the renin-angiotensin system (e.g.,
angiotensin-converting enzyme inhibitors, ACEIs) will precipitate acute renal
failure (HRICIK and DUNN 1990; WHELTON et al. 1990).
The role of prostaglandins in preserving glomerular filtration under condi-
tions of elevated circulating angiotensin II has recently been demonstrated in
man (MOTWANI and STRUTHERS 1994). In the clinical context, it should be
remembered that prolonged diuretic treatment stimulates the renin-angio-
tensin system, which in turn necessitates renal prostaglandins for preserved
glomerular filtration. In this situation, even the trivial administration of an
NSAID for minor musculoskeletal disorders may have serious consequences.
It is still not clear whether all NSAIDs are equi-effective in causing this
kind of nephrotoxicity or whether some of them, notably sulindac, may be
considered as safer drugs (WHELTON et al. 1990; JOHNSON et al. 1994).
Drug-Drug Interactions at Receptors and Other Active Sites 229
EFFERENT ARTERIOLE

PROSTAGLANOINS ANGIOTENSIN n

Fig. 5. Compensatory mechanisms to decreased renal perfusion and drug interactions.

Decreased renal perfusion due, for example, to intravascular volume depletion (diuret-
ics) or systemic hypotension (antihypertensive agents) results in decreased intra-
glomerular hydrostatic pressure and filtration rate. To compensate, both the
renin-angiotensin system and the production of prostaglandins are stimulated, mainly
in response to sympathetic system activation. Afferent arteriole vasodilation by pros-
taglandins and efferent arteriole vasoconstriction by angiotensin II restore glomerular
hydrostatic pressure and filtration rate. In this condition, drugs that interfere with
the compensatory mechanisms, such as the non-steroidal anti-inflammatory drugs
(NSAIDs) and angiotensin-converting enzyme inhibitors (ACEIs), may, either or both,
precipitate acute renal failure (see text for additional explanations)

2. Prostaglandins, Natriuresis and Antihypertensive Therapy

By interfering with the production of renal prostaglandins, the NSAIDs in-
duce hydro saline retention (see Sect. B.IV.1) and may reduce the natriuretic
effect of concurrently administered diuretics, particularly furosemide (FAVRE
et al. 1983). Moreover, the NSAIDs can decrease the antihypertensive effect
of most antihypertensive drugs (ZAHLER et al. 1990). This has been confirmed
in a recent meta-analysis of randomized trials studying the effect of NSAIDs
on blood pressure: among the NSAIDs, piroxicam had the greatest effect,
whereas sulindac and acetylsalicylic acid were the least effective (JOHNSON et
al. 1994). In view of the fact that some NSAID preparations have now become
available over the counter, the possibility of a drug interaction should be
considered when unexpected blood pressure elevation is detected in a previ-
ously stabilized hypertensive patient.

3. Potassinm Homeostasis
All thiazide and loop diuretics increase the renal excretion of potassium and
reduce kalaemia. Hypokalaemia augments the risk of arrhythmias, particu-
230 M. SCHORDERET and J.D. FERRERO

larly in the course of therapy with antiarrhythmic drugs (see Sect. B.I1.2). It is
also well known to increase the toxicity of cardiac glycosides (STEINESS and
OLESEN 1976). In order to treat hypokalaemia, certain patients will require the
prescription of potassium supplements or of a potassium-sparing diuretic. In
these patients, the co-prescription of ACE inhibitors creates a serious risk of
hyperkalaemia (BURNAKIS and MIODUCH 1984). By interfering with renin re-
lease, the NSAIDs tend to produce hyperkalaemia (TAN et al. 1979), which
would also be aggravated by the simultaneous prescription of potassium-
sparing diuretics or of potassium salts (ZIMRAN et al. 1985).

References
Amrein R, Giintert TW, Dingemanse J, Lorscheid T, Stabl M, Schmid-Burgk W (1992)
Interactions of moclobemide with concomitantly administered medication: evi-
dence from pharmacological and clinical studies. Psychopharmacology (Berl)
106:S24-S31
Angelo-Nielson K (1980) Timolol topically and diabetes mellitus. JAMA 244:263
Balfour JA, McTavish D (1993) Acarbose. An update of its pharmacology and
therapeutic use in diabetes mellitus. Drugs 46:1025-1054
Bauman JL, Grawe JJ, Winecoff AP, Hariman RJ (1994) Cocaine-related sudden
cardiac death: a hypothesis correlating basic science and clinical observations. J
Clin Pharmacol 34:902-911
Ben-David J, Zipes DP (1993) Torsades de pointes and proarrhythmia. Lancet
341:1578-1582
Berlin I, Zimmer R, Cournot A, Payan C, Peclariosse AM, Puech AJ (1989) Determi-
nation and comparison of the pressor effect of tyramine during long-term
moclobemide and tranylcypromine treatment in healthy volunteers. Clin
Pharmacol Ther 46:344-351
Bigger JT (1984) The interaction of mexiletine with other cardiovascular drugs. Am
Heart J 107:1079-1085
Blackwell B, Marley E (1966) Interactions of cheese and its constituents with monoam-
ine oxidase inhibitors. Br J Pharmacol Chemother 26:120-141
Blakely RD, Berson HE, Fremeau RT, Caron MG, Peek MM et al. (1991) Cloning
and expression of a functional serotonin transporter from rat brain. Nature
354:66-70
Bochner BS, Lichstenstein LM (1991) Anaphylaxis. N Engl J Med 324:1785-1790
Boess FG, Martin IL (1994) Molecular biology of 5-HT receptors. Neuropharmacology
33:275-317
Browne B, Linter S (1987) Monoamine oxidase inhibitors and narcotic analgesics. Br J
Psychiatry 151:210-212
Bugnon 0, Dommer J, Mesnil M, Schorderet M (1993) Detection and management of
drug interactions (in French). Swiss Society of Pharmacy, Bern (Basic formation,
vol 1)
Burnakis TG, Mioduch HJ (1984) Combined therapy with captopril and potassium
supplementation. A potential for hyperkalemia. Arch Intern Med 144:2371-2372
Butler WT, Rossen RD (1977) Effects of corticosteroids on immunity in man. I.
Decreased serum IgG concentration caused by 3 or 5 days of high doses of
methylprednisolone. J Clin Invest 30:1451-1455
Cetaruk EW, Aaron CK (1994) Hazards of nonprescription medications. Emerg Med
Clin North Am 12:483-510
Christ W, Lehnert T, Ulbrich B (1988) Specific toxicologic aspects of the quinolones.
Rev Infect Dis 10 [Suppl1]:S141-S146
Clarke RSJ, Mirakhur RK (1994) Adverse effects of muscle relaxants. Adv Drug React
Toxicol Rev 13:23-41
Drug-Drug Interactions at Receptors and Other Active Sites 231

Cooper JR, Bloom FE, Roth RH (1991) The biochemical basis of neuropharmacology,
6th edn. Oxford University Press, New York
Curtis DR, Game CJA, Johnston GAR, McCulloch RM, MacLachlan RM (1972)
Convulsive action of penicillin. Brain Res 43:242-245
Dilsaver SC (1993) Antipsychotic agents: a review. Am Fam Physician 47:199-
204
Drachman DB (1994) Myasthenia gravis. N Engl J Med 330:1797-1809
Drug and Therapeutics Bulletin (1992) Sumatriptan: a new approach to migraine. Drug
Ther Bull 30:85-87
Edwards G, Weston AH (1993) The pharmacology of ATP-sensitive potassium chan-
nels. Annu Rev Pharmacol Toxicol 33:597-637
Favre L, Glasson P, Riondel A, Vallotton MB (1983) Interaction of diuretics and non-
steroidal anti-inflammatory drugs in man. Clin Sci 64:407-415
Feighner JP, Boyer WF, Tyler DL, Neborsky RJ (1990) Adverse consequences of
fluoxetine-MAOI combination therapy. J Clin Psychiatry 51:222-225
Fiore MC, Jorenby DE, Baker TB, Kenford SL (1992) Tobacco dependence and the
nicotine patch; clinical guidelines for effective use. JAMA 268:2687-2694
Fitton A, Faulds D, Goa KL (1992) Moclobemide. A review of its pharmacological
properties and therapeutic use in depressive illness. Drugs 43:561-596
Gerich JE (1989) Oral hypoglycemic agents. N Engl J Med 321:1231-1245
Giardia E, Weschsler M (1990) Low dose quinidine-mexiletine combination therapy
versus quinidine monotherapy for treatment of ventricular arrhythmias. J Am Coli
Cardiol 15:1138-1145
Gieschke R, Schmid-Burgk W, Amrein R (1988) Interaction of moclobemide, a new
reversible monoamine oxidase inhibitor with oral tyramine. J Neural Transm
[Suppl 26]:97-104
Giros B, Mestikawy L, Bertrand L, Caron MG (1991) Cloning and functional
characterization of a cocaine-sensitive dopamine transporter. FEBS Lett 295:149-
154
Glassman AH, Bigger JT Jr (1981) Cardiovascular effects of therapeutic doses of
tricyclic antidepressants. Arch Gen Psychiatry 38:815-820
Gram LF (1994) Fluoxetine. N Engl J Med 331:1354-1361
Gross P A, Gould A, Brown AE (1985) Effect of cancer chemotherapy on the immune
response to influenza virus vaccine: review of published studies. Rev Infect Dis
7:613-618
Guy EJ (1992) Selective serotonin reuptake inhibitors. Br Med J 304:1655-1658
Halliwell RF, Davey PG, Lambert 11 (1993) Antagonism of GABAA receptors by 4-
quinolones. J Antimicrob Chemother 31:457-462
Hansten PD, Horn JR, Koda-Kimble MA, Young LY (eds) (1993) Drug interactions
and updates quarterly. Applied Therapeutics, Vancouver
Hill S, Yau K, Whitwam J (1992) MAOIs to RIMAs in anaesthesia - a literature
review. Psychopharmacology (Berl) 106 [Suppl]:S43-S45
Hoffman BJ, Mezby E, Brownstein MJ (1991) Cloning of a serotonin transporter
affected by antidepressants. Science 254:79-80
Horn PT, Murphy MB (1991) Dopamine receptor agonists in cardiovascular medicine.
Trends Cardiovasc Med 1:103-107
Hricik DE, Dunn MJ (1990) Angiotensin-converting enzyme inhibitor-induced
renal failure: causes, consequences, and diagnostic uses. J Am Soc Nephrol1:845-
858
Jackson EJ, Bressler R (1981) Clinical pharmacology of sulphonylurea hypoglycemic
agents. Drugs 22:211-245, 295-320
Johnson AG, Nguyen TV, Day RO (1994) Do nonsteroidal anti-inflammatory drugs
affect blood pressure? A meta-analysis. Ann Intern Med 121:289-300
Kilty JE, Lorang D, Amara SG (1991) Cloning and expression of a cocaine-sensitive rat
dopamine transporter. Science 254:578-579
Kotler MN, Berman L, Rubenstein AH (1966) Hypoglycaemia precipitated by pro-
pranolol. Lancet 11:1389-1390
232 M. SCHORDERET and J.D. FERRERO

Lange RA, Cigarro RG, Flores ED, McBride W, Kim AS, Bedotto JB, Danziger RS,
Hillis LD (1990) Potentiation of cocaine-induced coronary vasoconstriction by
beta-adrenergic blockade. Ann Intern Med 324:1485-1490
Langieri G, Marinchak RA, Rials SJ, Kowey PR (1995) Antiarrhythmic drug interac-
tions. Curr Op in Cardiol 10:26-28
Lejoyeux M, Ades J, Rouillon F (1994) Serotonin syndrome. Incidence, symptoms and
treatment. CNS Drugs 2:132-143
Lewitt MS, Yu VK, Rennie GC, Carter IN, Marel GM, Yue DK, Hooper MJ (1989)
Effects of combined insulin-sulfonylurea therapy in type II patients. Diabetes
Care 12:379-383
Marchant B, Wray R, Leach A, Nama M (1989) Flumazenil causing convulsions and
ventricular tachycardia. Br Med J 299:860
Mars H (1974) Levodopa, carbidopa and pyridoxine in Parkinson's disease. Arch
NeuroI30:444-447
McInnes GT, Brodie MJ (1988) Drug interactions that matter. A critical reappraisal.
Drugs 36:83-110
Mefford IN (1988) Epinephrine in mammalian brain. Prog Neuropsychopharmacol
Bioi Psychiatry 12:365-388
Michell JR, Cavanaugh JH, Arias L, Oates JA (1970) Guanethidine and related agents.
III. Antagonism by drugs which inhibit the norepinephrine pump in man. J Clin
Invest 49:1596-1604
Micossi P, Pontiroli AE, Baron SH, Tamayo RC, Lengel F, Bevilacqua M, Raggi U,
Norbiato G, Foa PP (1978) Aspirin stimulates insulin and glucagon secretion and
increases glucose tolerance in normal and diabetic subjects. Diabetes 27:1196-
1204
Mitchell PB (1994) Selective serotonin reuptake inhibitors: adverse effects, toxicity and
interactions. Adv Drug React Toxicol Rev 13:121-144
Mitchelson F (1992) Pharmacological agents affecting emesis; a review (part I). Drugs
43:295-315
Mody I, De Koninck Y, Otis TS, Soltesz I (1994) Bridging the cleft at GABA synapses
in the brain. Trends Neurosci 17:517-525
Mott DD, Lewis DV (1994) The pharmacology and function of central GABAB recep-
tors. Int Rev Neurobiol 36:97-223
Motwani JG, Struthers AD (1994) Interactive effects of indomethacin, angiotensin II
and frusemide on renal haemodynamics and natriuresis in man. Br J Clin
PharmacoI37:355-361
Neuvonen PJ, Pohjola-Sintonen S, Tacke U (1993) Five fatal cases of serotonin syn-
drome after moclobemide-citalopram or moclobemide-clomipramine overdoses.
Lancet 342:1419
Pacholczyk T, Blakely RD, Amara SG (1991) Expression cloning of a cocaine-
and antidepressant-sensitive human noradrenaline transporter. Nature 350:350-
353
Pandit MK, Burke J, Gustafson AB, Minocha A, Peiris AN (1993) Drug-induced
disorders of glucose tolerance. Ann Intern Med 118:529-539
Remick RA (1988) Anticholinergic side effects of tricyclic antidepressants and
their management. Prog Neuropsychopharmacol Bioi Psychiatry 12:225-
231
Richardson T, Foster J, Mawer GE (1986) Enhancement by sodium salicylate of the
blood glucose lowering effect of chlorpropamide-drug interaction or summation of
similar effects? Br J Clin Pharmacol 22:43-48
Schmid-Antomarchi H, De Weille J, Fosset M, Lazdunski M (1987) The receptor for
antidiabetic sulfonylureas controls the activity of the ATP-modulated K+ channel
in insulin-secreting cells. J Bioi Chern 262:15840-15844
Schou JS (1982) Drug interactions at (pharmacodynamically active) receptor sites.
Pharmacol Ther 17:199-210
Schourkes JA, Van Bergen PJ, Van Nueten JM (1988) Prejunctional muscarinic (M)-
Drug-Drug Interactions at Receptors and Other Active Sites 233

receptor interactions on guinea-pig ileum: lack of effect of cisapride. Br J

Pharmacol 94:228-234
Seeman P, Van Tol HM (1994) Dopamine receptor pharmacology. Trends Pharmacol
Sci 15:264-270
Seltzer HS (1989) Drug-induced hypoglycemia. A review of 1418 cases. Endocrinol
Metab Clin North Am 18:163-183
Shimada S, Kitamaya SK, Lin CL, Patel A, Nanthakumar P, Gregor P, Kuhar M, Uhl
G (1991) Cloning and expression of a cocaine-sensitive dopamine transporter
complementary DNA. Science 254:576-578
Sieghart W (1992) GABAA receptors: ligand-gated Cl- ion channels modulated by
multiple drug-binding sites. Trends Pharmacol Sci 13:446-450
Singh YN, Marshall IG, Harvey AL (1982) Pre- and postjunctional blocking effects of
aminoglycosides, polymyxin, tetracycline and lincosamide antibiotics. Br J
Anaesth 54:1295-1305
Siscovick DS, Raghunathan TE, Psaty BM, Koepsell TD, Wicklund KG, Lin X, Cobb
L, Rautaharju PM, Cop ass MK, Wagner EH (1994) Diuretic therapy for hyperten-
sion and the risk of primary cardiac arrest. N Engl J Med 330:1852-1857
Spina E, Perucca E (1994) Newer and older antidepressants. A comparative review of
drug interactions. CNS Drugs 2:479-497
Stack CG, Rogers P, Linter SPK (1988) Monoamine oxidase inhibitors and
anaesthesia. A review. Br J Anaesth 60:222-227
Steiness E, Olesen KH (1976) Cardiac arrhythmias induced by hypokalaemia and
potassium loss during maintenance digoxin therapy. Br Heart J 38:167-172
Stockley IH (1991) Drug interactions. Blackwell, Oxford
Sugrue MF (1981) Current concepts on the mechanisms of action of antidepressant
drugs. Pharmacol Ther 13:219-247
Tan SY, Shapiro R, Franco R, Stockard H, Mulrow PJ (1979) Indomethacin-induced
prostaglandin inhibition with hyperkalemia. A reversible cause of hyporeninemic
hypo aldosteronism. Ann Intern Med 90:783-785
Usdin TB, Mezey E, Chen C, Brownstein MJ, Hoffman BJ (1991) Cloning of the
cocaine-sensitive bovine dopamine transporter. Proc Natl Acad Sci USA
88:11168-11171
Whelton A, Stout RL, Spilman PS, Klassen DK (1990) Renal effects of ibuprofen,
piroxicam, and sulindac in patients with asymptomatic renal failure. Ann Intern
Med 112:568-576
Whitcomb DC, Gilliam FR III, Starmer CF, Grant AO (1989) Marked QRS complex
abnormalities and sodium channel blockade by propoxyphene reversed with
lidocaine. J Clin Invest 84:1629-1636
White NJ (1994) Mefloquine in the prophylaxis and treatment of falciparum malaria.
Br Med J 308:286-287
Winniford MD, Markham RV Jr, Firth BG, Nicod P, Hillis LD (1982) Hemodynamic
and electrophysiologic effects of verapamil and nifedipine in patients on
propranolol. Am J CardioI50:704-710
Woosley RL (1991) Antiarrhythmic drugs. Annu Rev Pharmacol Toxicol 31:427-455
Zahler SM, al-Kiek R, Ivanovich P, Mujais SK (1990) Nonsteroidal anti-inflammatory
drugs and antihypertensives: cooperative malfeasance. Nephron 56:345-352
Zielenski CC, Stuller I, Dorner F, Potzi P, Muller C, Eibl MM (1986) Impaired
primary, but not secondary, immune response in breast cancer patients under
adjuvant chemotherapy. Cancer 58:1648-1652
Zimran A, Kramer M, Plaskin M, Herschko C (1985) Incidence of hyperkalaemia
induced by indomethacin in a hospital popUlation. Br Med J 291:107-108
CHAPTER 8
Synergistic Drug Interactions
J.P. GRIFFIN and P.F. D' ARCY

A. Introduction: "Mithridatium"
Synergy is derived from the Greek word O'Vv£P}f5s, meaning working together.
In pharmacological terms it has a precise meaning, which is distinct from
interaction between two or more active substances where the interaction may
be on the absorption, distribution, metabolism or excretion of one or more of
the substances. For a synergistic therapeutic effect to occur there need not
necessarily be an interaction of this type but the effect should be to the
patient's therapeutic advantage.
Historically, the seeking for the universal panacea or cure for all ills has
to some extent hinged on the philosophy of empirical polypharmacy. This
can be traced back to philosophies such as that evolved by Mithridates VI,
King of Pontus 133 B.C.-63 B.c. (see GRIFFIN 1994). Pontus abounded in
medicinal plants and Mithridates acquired considerable knowledge of
them. Like every despot of that period he lived in fear of being assassinated
by poisoning, in consequence of which he sought the universal antidote to
all poisons. Mithridates proceeded along a simple line of reasoning. Having
investigated the powers of a number of single ingredients, which he found
to be the antidote to various venoms and poisons individually, he evaluated
them experimentally on condemned criminals. He then compounded all the
effective substances into one antidote hoping thereby to produce universal
protection. A daily dose was taken prophylactically to give the immunity he
sought.
Pliny wrote: "by his unaided efforts Mithridates devised the plan of drink-
ing poison daily after first taking remedies in order to achieve immunity by
sheer habituation. He was the first to discover the various antidotes, one of
which is even known by his name". So effective was Mithridates' formulation
that he tried unsuccessfully to commit suicide by poisoning and finally killed
himself with a "Celtic sword".
Galen writing in the second century A.D. at a time when he was physician
to the Roman Emperor, Marcus Aurelius, refers to "Mithridatium" and a
formulation derived from it by one Andromachus, Nero's physician. It is said
that Andromachus removed some ingredients from Mithridates' formulation
and added others particularly viper's flesh. To this new product he gave the
name "Galene", which means "tranquillity". Galene later became known as
236 J.P. GRIFFIN and P.F. D' ARCY

theriac. Details of various theriacs, including Mithridatium and Galene, were

given in Galen's Antidotes I and Antidotes II.
In Galen's Antidotes I he distinguished three kinds of antidote, those
that countered poisons, those that countered venoms and those that count-
ered ailments. Some would counter all three and Galen claimed that
Mithridatium and Galene belonged to this latter class. According to Galen,
Mithridatium contained 41 ingredients and the Galene of Andromachus 55
components.
The use of Mithridatium persisted in Britain until the 1746 London
Pharmacopoeia; it was removed from the 1788 edition. The Edinburgh
Pharmacopoeia included it in its 1699 edition but dropped it from the 1759
edition. However, Mithridatium persisted in the German Pharmacopoeia of
1872 and the French Pharmacopoeia of 1884.
WILLIAM HEBERDEN in 1745 described Mithridatium and theriacs as: "this
medley of discordant samples ... made up of a dissonant crowd collected from
many countries, mighty in appearance, but in reality, an ineffective multitude
that only hinder one another".
What was written by WILLIAM HEBERDEN two and a half centuries ago
could equally well be applied to a considerable number of the multi-ingredient
products that persist in use today, particularly in many of the traditional OTC
remedies and in many of the unlicensed Indian and Chinese medicines which,
although unlicensed, are freely available throughout Europe through various
ethnic outlets.
In current medical practice a number of examples of co-formulations and/
or co-prescribing of active ingredients on the basis of a believed synergistic
action or interaction can be cited. For many of these the evidence for the
combined use has been derived on the basis of theory which mayor may not
stand up to critical analysis in terms of practical therapeutics.
A brief account of examples of the more important treatments with be-
lieved synergistic drug combinations is presented in this chapter. The ex-
amples start with the early use of probenecid to slow the renal elimination of
penicillin. They follow on with the demonstration of synergistic drug effects in
the use of diuretics, with the synergistic design of co-trimoxazole, with syner-
gism in the use of antiepileptics, antitubercular agents, antileprosy drugs, and
drugs to treat peptic and duodenal ulcers, and finally with synergism in the
control of diabetes and in cancer chemotherapy.

B. Early Use of Probenecid with Penicillin

Many acidic drugs and drug metabolites are actively secreted by the proximal
renal tubular active transport mechanism and interactions may arise from
competition for this system. Drugs actively secreted include the penicillins and
probenecid; thus the plasma half-life of penicillin may be prolonged by
probenecid. The newer penicillins have rendered this combination obsolete
Synergistic Drug Interactions 237

although the inhibition of the urinary excretion of penicillin and cephalospor-

ins has been used as a device to increase the biliary excretion of these agents,
thereby raising the antibiotic concentrations in the biliary tract. This has been
used to improve the efficacy of antibiotic treatment of cholecystitis (SALES
et al. 1972) (see also, Chap. 6, this volume).

C. Diuretic Combinations
Commonly prescribed fixed combinations that can be cited as examples of
believed benefit are combined diuretics. The BRITISH NATIONAL FORMULARY
No. 27 (1994) cites over 25 products comprising a loop or thiazide diuretic
combined with either potassium or a potassium-sparing diuretic. The MeRec
Bulletin (MEDICINES RESOURCE CENTRE 1994) asks the question: Are such
combinations necessary? and concludes that for the vast majority of patients
there is no need for such combinations, and that the routine approach should
be to give a thiazide or loop diuretic alone.

D. Co-trimoxazole
In 1948, HITCHINGS et al. demonstrated that nearly all 2,4-diaminopyrimidines
and related compounds inhibited the growth of Lactobacillus casei by virtue of
their properties as folic acid antagonists. Four years later the precise mode of
action was shown to involve inhibition of the enzyme dihydrofolate reductase.
This is the enzyme which catalyses the transformation of folic acid into a form
that can be utilised in the process leading to thymine formation prior to its
incorporation into DNA. It was eventually shown that the degree of inhibition
of the enzyme by different antimetabolites varied according to the species
from which the enzyme was isolated. This led to the development of species-
specific dihydrofolate reductase inhibitors amongst a series of 5-benzyl deriva-
tives of 2,4-diaminopyrimidine. The antimalarial drug pyrimethamine is one
such example. By varying the substituents on the benzene and pyrimidine
rings, it was found that methoxyl groups enhanced the inhibitory potency
against cultures of Proteus vulgaris, whilst not markedly inhibiting the enzyme
derived from mammalian sources.
This discovery was fully exploited with trimethoprim, the trimethoxy
derivative, which is 50000 times more potent as an inhibitor of bacterial
dihydrofolate reductase as it is of human dihydrofolate reductase (see
SNEADER 1985). Trimethoprim was marketed in fixed combination with
sulphamethoxazole as Septrin and Bactrim by Wellcome and Roche, respec-
tively. This combination was based on the fact that sulphonamides prevent the
bacterial synthesis of dihydrofolate. It was theoretically conceived that if a
bacterium was resistant to the sulphonamide it could be hoped that the
dihydrofolate acid reduction process would be inhibited by trimethoprim or
vice versa. The choice of sulphamethoxazole as the sulphonamide in the
238 J.P. GRIFFIN and P.F. D'ARCY

combination was based purely on the fact that it had approximately the same
duration of action as trimethoprim.
The eventual choice of the name co-trimoxazole by the British
Pharmacopoeia Commission gave a form of sanctity to this combination and it
was a sanctity that time has shown to be ill-deserved.
The BRITISH NATIONAL FORMULARY No. 27 states that "trimethoprim alone
can be used for the treatment of urinary and respiratory-tract infections and
for prostatitis, shigellosis and invasive salmonella infections. Side effects are
less than with co-trimoxazole especially in older patients; therefore it should
be used in place of co-trimoxazole for most infections (except Pneumocystis
carinii infections and respiratory infections associated with AIDS)".
The side effects of co-trimoxazole are the same as those of sulphonamides,
e.g., rashes, Stevens-Johnson syndrome, toxic epidermal necrolysis, renal
failure and blood dyscrasias notably bone-marrow depression and
agranulocytosis. In short, the serious side effects of co-trimoxazole are due to
the sulphonamide component which, apart from a possible but dubious syner-
gistic role in Pneumocystis carinii infections, is not serving any useful purpose.
This example of co-trimoxazole is one where a synergistic role was con-
ceived on good theoretical grounds but was shown after some 12 years to be
unnecessary and undesirable and unsafe for many patients, particularly the
elderly. The sulphonamide component contributes little to the efficacy of the
product but causes most of the combination's adverse reactions.
Despite the fact that the undesirability of this fixed combination has been
known for over 15 years and that repeated comments to this effect have been
made in the BRITISH NATIONAL FORMULARY in every issue since 1981, the
product remains widely used by British general practitioners. Some 18 million
prescriptions per year are written in general practice for co-trimoxazole,
and approximately the same number for trimethoprim. However, in hospital
practice, the use of co-trimoxazole has virtually disappeared.

E. Combination Treatment in Control of Epilepsy

For many decades epileptic patients were treated with combinations of
anticonvulsant drugs, one of the more common combinations being
phenobarbitone and phenytoin. This combination is still one of the most
widely used medications for epileptic patients. GRIFFIN and WYLES (1991) state
that "until the 1970s poly therapy (more than one anticonvulsant) was the most
widely used and accepted practice. Monotherapy is now generally recognised
as being superior. The basis of mono therapy is accurate diagnosis followed by
the selection of a single appropriate anticonvulsant".
Table 1 shows the drugs of choice for epilepsy (after CHADWICK 1993).
The BRITISH NATIONAL FORMULARY No. 1 in 1981 stated "therapy with
several drugs at once should be avoided. Patients are best controlled with one
anti epileptic. Combinations of drugs often used to be given on the grounds
Synergistic Drug Interactions 239

Table 1. Antiepileptic drugs: single therapy by epilepsy type. (After CHADWICK 1993)

Epilepsy type First choice Second choice

Idiopathic
Simple absences Sodium valproate or Benzodiazepines or
ethosuximide lamotrigine
Juvenile myoclonic Sodium valproate Phenobarbitone or
lamotrigine
Awakening time/clonic Sodium valproate Carbamazepine, phenytoin
or lamotrigine
Symptomatic Sodium valproate or Carbamazepine, phenytoin
benzodiazipines or phenobarbitone
Partial epilepsy Carbamazepine or Phenytoin, phenobarbitone
sodium valproate or vigabatrin
Unclassified Sodium val pro ate Lamotrigine

that their therapeutic effects were additive while their individual toxicity was
reduced, but there is no evidence for this." The wording in the BNF No. 27
(1994) is identical.
The continuing high usage of the phenobarbitone/phenytoin combinations
shows how hard it is to break an established but undesirable prescribing
practice. The usage of the new anticonvulsants (e.g., GABA derivatives,
vigabatrin) remains low. The popularity of the ingrained phenobarbitone/
phenytoin combination is, however, made more attractive by the annual low
price for treatment (see Table 2). Downward pressures by Government on
prescribing costs are an undoubted factor in perpetuating the use of two agents
once thought to have a useful synergistic action. Conversely, the use of new
single-therapy regimens is being discouraged by cost.

F. Antitubercular Drugs
The synergistic use of antibiotics and antibacterial agents in the treatment of
tuberculosis is well established. The treatment is in two phases, an "initial
phase" with at least three drugs and a "continuation phase" using two drugs.
The following regimen is recommended by the British Thoracic Society:
Initial Phase. The concurrent use of three antitubercular agents is designed to
reduce the population of viable bacteria as rapidly as possible. The usual
regimen is the daily use of rifampicin, isoniazid and pyrazinamide; ethambutol
may be added if resistant organisms are suspected. Streptomycin is rarely used
in the United Kingdom, although it is an effective agent for use in the "initial
phase" and can be used instead of isoniazid. The "initial phase" of treatment
should last at least 2 months.
Continuous Phase. This should last a minimum of 4 months using a combina-
tion of isoniazid and rifampicin.
240 J.P. GRIFFIN and P.F. D' ARCY

Table 2. Comparison of daily dose and annual cost of

anticonvulsants. (From CRIFFIN and WYLES 1991)

Anticonvulsant Annual cost Commonest

daily dose

Carbamazepine £65.00 600mg

Clonazepam £101.00 6mg
Ethosuximide £94.00 1000mg
Phenytoin £28.00 300mg
Phenobarbitone £13.00 180mg
Primidone £27.00 1000mg
Sodium valproate £96.00 800mg
Vigabatrin £168.00 3000mg

Treatment regimens for Third World countries may well differ from that
recommended in the United Kingdom and will largely depend on the local
pattern of resistance and the cost and availability of antitubercular drugs.
Dosage regimens can be found in the BRITISH NATIONAL FORMULARY and
the current edition should always be consulted.

G. Anti-Leprosy Treatments
For more than 20 years the use of dapsone alone was commonplace but
multidrug regimens have had to be developed due to the emergence of drug-
resistant strains of the leprosy bacillus. Thus, the principle for the use of multi-
drug treatments in leprosy is the same as that of tuberculosis. For therapeutic
purposes, leprosy can be subdivided into, firstly, multibacillary leprosy, for
which a three-drug regimen is recommended using dapsone, rifampicin and
clofazimine; secondly, paucibacillary leprosy, for which a two-drug regimen of
rifampicin and dapsone are recommended. A 2-year period of treatment is
recommended in both grades of severity and advice from the Panel of Leprosy
Opinion from the Department of Health should be sought before initiating
treatment.

H. Peptic Ulcer Therapy

For many years the treatment of gastric and duodenal ulcer was based on the
use of antacids such as aluminium hydroxide and magnesium hydroxide. This
was supplemented by bed rest and milk alkali drip feeds through a nasogastric
tube in conjunction with aspiration of acid. Surgical intervention with partial
gastrectomy, Polya gastrectomy, Billroth gastrectomy or vagotomy and
pyloroplasty were common until the introduction of the Hz-antagonist
cimetidine in the late 1970s, shortly followed by other Hz-antagonists
ranitidine, famotidine and nizatidine in the early 1980s. Misoprostol, a syn-
thetic prostaglandin analogue, and proton pump inhibitors such as omeprazole
Synergistic Drug Interactions 241

revolutionised the treatment of peptic ulcer and virtually eliminated the need
for the heroic gastric surgery conducted in earlier decades.
Treatment of peptic ulcer is due to undergo a further revolution with
the discovery of Helicobacter pylorii as the causative organism. In a press
release following the VIIth Workshop on Gastroduodenal Pathology and
Helicobacter pylorii held in Texas in September 1994, the detailed results of a
trial by Dr. R.P.H. Logan of The Queen's Medical Centre, Nottingham, UK,
were presented in detail; they are given below:
Dr. Logan presented I-year follow-up data on a multicenter study of 154
patients with duodenal ulcers and H. pylorii infection. Six months data from
the same study were presented in May during Digestive Disease Week in New
Orleans.
Participants in the trial received either 500 mg clarithromycin (three times
a day) or placebo along with 40mg omeprazole (once a day) during the first 2
weeks of the study; all study participants received 40 mg omeprazole for an
additional 2 weeks.
According to Dr. Logan's 6-months interim results, ulcers healed after 4
weeks in 100% (64/64) of patients receiving clarithromycin and omeprazole
and 99% (71/72) of patients receiving omeprazole and placebo. However, a
significant difference was observed between the two study arms in the eradica-
tion of H. pylorii at 4-6 weeks after the conclusion of therapy. Patients who
were treated with the clarithromycin/omeprazole combination showed H.
pylorii rates of 83% - or 50 out of 60 evaluable patients - at 4-6 weeks post-
treatment versus 1 % (1174) in the group receiving placebo/omeprazole.
Furthermore, 6 months after treatment only 4% (2/53) of patients who
received clarithromycin and omeprazole experienced ulcer recurrence, versus
55% (36/60) of those who received placebo and omeprazole.
One year after treatment, the final results indicate incidence of ulcer
recurrence was 6% (3/51) in patients treated with clarithromycin and
omeprazole, versus 76% (47/62) of patients who received a placebo
and omeprazole. These data suggested that dual therapy with clarithromycin
and omeprazole might provide an effective regimen to heal and prevent occur-
rence of ulcers (LOGAN 1994).
The clarithromycin and omeprazole combination appeared to be well
tolerated according to researchers. Dr. Logan reported that only four patients
receiving clarithromycin and omeprazole and one patient receiving
omeprazole with placebo discontinued therapy because of adverse events.
Adverse events that occurred more frequently in the clarithromycinl
omeprazole-treated patients were tongue discoloration and taste perversion.
In this study, no patient experienced a serious adverse event.
At the Xth World Congress of Gastroenterology held in Los Angeles (2-
7 October 1994), a paper by GUSTAVSON et al. was presented. This dealt with
the pharmacokinetic interactions between omeprazole and clarithromycin in
20 male healthy volunteers. The results of their study led the authors to draw
three salient conclusions, namely:
242 J.P. GRIFFIN and P.F. D'ARCY

1. Administration of omeprazole with clarithromycin resulted in higher and

more prolonged plasma concentrations of omeprazole than when
omeprazole was administered alone, but there was no significant difference
in the effect of omeprazole on gastric pH.
2. Administration of clarithromycin with omeprazole resulted in somewhat
higher plasma concentrations of clarithromycin and 14(R)-hydroxy-
clarithromycin than when clarithromycin was administered with placebo.
3. Good penetration of clarithromycin and 14(R)-hydroxy-clarithromycin
into gastric tissues and mucus was observed. Tissue and mucus concentra-
tions were higher when clarithromycin was administered with omeprazole
than when clarithromycin was administered with placebo.
Other regimens combining an acid control and antibiotic element are cur-
rently under clinical evaluation, for example:
1. Ranitidine bismuth citrate with clarithromycin
2. Omeprazole plus amoxycillin plus metronidazole
3. Omeprazole plus amoxycillin plus clarithromycin
4. Bismuth plus amoxycillin plus metronidazole
In addition experimental work from Japan suggested that degrading the cell
wall of Helicobacter pylorii with a proteolytic enzyme in combination with the
proton pump inhibitor lansoprazole in combination with bismuth, amoxycillin
and metronidazole indicated that there might be an improved antibacterial
effect with the use of pronase.
The work on H. pylorii elimination by the synergistic action of various
combinations indicated that the clinical evaluation of synergistic actions
should and could be evaluated. The "armchair" combination of treatments
based on sound theoretical considerations are hopefully a thing of the past.
The first combination approved both in the United Kingdom and Sweden for
H. pylorii infections is omeprazole and amoxycillin. It would be reasonable to
expect that regulatory approval for recommendation to use omeprazole and
clarithromycin in combination can be expected shortly.
Most clinical researchers now believe that adding another antibacterial to
the currently accepted dual therapy would improve the H. pylorii cure rate;
metronidazole for 1 week appears to be the frontrunner for this role.
A paper from Hong Kong by HOSKINGS et al. (1994) demonstrated that a
I-week cource of bismuth, tetracycline and metronidazole in patients with
duodenal ulcer and H. pylorii infection led to ulcer healing in 92 % of patients.
The addition of omeprazole in a similar-sized group of patients led to ulcer
healing in 95%. The authors of this report concluded that the addition of
omeprazole greatly reduced ulcer pain but had no effect on ulcer healing.

I. Non-Insulin-Dependent (Maturity Onset) Diabetes

Oral antidiabetic drugs are used in the control of non-insulin-dependent (type
2, obesity or maturity onset) diabetes. They should not be prescribed until
Synergistic Drug Interactions 243

patients have been shown not to respond adequately to 3 months restriction of

calorie and carbohydrate intake. They should be used to augment attempts to
control the diabetes by diet.
There are two classes of oral hypoglycaemics: firstly, the sulponylureas,
which act mainly by augmenting insulin secretion and consequently depend on
the patient having some remaining pancreatic f3-cell function. In long-term
therapy there is also thought to be an extrapancreatic action. The second
group are the biguanides; metformin is the only available biguanide currently
on the British market; phenformin was withdrawn due to its propensity to
cause lactic acidosis. Metformin can also induce lactic acidosis but this occurs
almost exclusively in patients with renal impairment, in whom it is contra-
indicated. Metformin exerts its effect by a different mode of action from the
sulphonylureas, mainly by decreasing gluconeogenesis and by increasing the
peripheral utilisation of glucose. Since it acts only in the presence of insulin
it is effective only in those diabetics with some functional pancreatic islet B
cells.
In 1970, the UNIVERSITY GROUP DIABETES PROGRAM (UGDP) in the United
States described the results from the first 8112 years of study in 12 centres
in non-insulin-requiring diabetics. These patients were randomly allocated
to five treatment groups: placebo; a standard dose of the sulphonylurea
tolbutamide; a standard dose of insulin; a variable dose of insulin; or a stan-
dard dose of phenformin. The study findings indicated that the combination of
diet plus tolbutamide was no more effective than diet alone in prolonging life.
Moreover, it was suggested that tolbutamide and diet may be less effective
than diet alone or diet with insulin as far as cardiovascular mortality was
concerned. A year later a similar conclusion was reached regarding
phenformin (KNATTERUD et al. 1971).
PROUT (1975) demonstrated in his analysis that diabetic patients under
active therapy with tolbutamide at the time of myocardial infarction were less
likely to survive than if no sulphonylurea was being administered.
SHANKS (1979) held that these observations suggested that the increased
mortality in patients receiving tolbutamide was directly related to an effect of
the drug rather than to effects which might result, for example, on unrestricted
diet. The mechanism by which tolbutamide might influence the outcome of an
acute myocardial infarction was unclear.
A retrospective study was carried out in Birmingham in 1974 of 184 known
diabetics who sustained a myocarial infarction (SOLER et al. 1974). The mortal-
ity rate was higher in patients treated with insulin or oral hypoglycaemic
agents than in non-diabetics, but earlier deaths were commoner in the patients
on oral therapy and these patients had a higher incidence of primary ventricu-
lar fibrillation.
Such effects may account for the increased mortality among patients
treated with tolbutamide who develop myocardial infarction. The factors
which increase the incidence of primary ventricular fibrillation in patients
receiving tolbutamide are not known although in vitro studies have shown that
244 J.P. GRIFFIN and P.F. D'ARCY

sulphonylureas have an ionotropic effect on heart muscle and increase the

automaticity of Purkinje fibres; these factors may increase the extent of myo-
cardial damage and lead to arrhythmias (LASSETER et al. 1972). It is clear that
this subject required further investigation.
It should be borne in mind that the UGDP study also showed an increased
mortality with the biguanide phenformin. Despite the recommendations of
SHANKS (1979), the authors know of no further investigations to shed light on
the clinical observations of the UGDP study. In fact the findings of the UGDP
study appear to have been forgotten over the last 2 decades and the only real
effect on practice has been to try and control maturity onset diabetic patients
with diet alone and only after this has failed to resort to the use of a
sulphonylurea. If diet plus a sulphonylurea fails to achieve control of the
diabetic then metformin may be added to the regimen (ALBERTI and
HOCKADAY 1987; BRITISH NATIONAL FORMULARY 1994).
ALBERTI and HOCKADAY (1987) state that "the biguanide's principal role
in treatment is as a synergist to a sulphonylurea when this together
with dietary advice has failed to produce satisfactory glucose control".
These authors then go on to state that "failure is often an indication to transfer
to insulin". Perhaps this would, in general, be the preferred course of
treatment.
More than 2 decades ago, a new area of research in diabetes involving
aldose reductase inhibitors (ARls) was introduced. They were a new class,
of drugs aimed at the consequences of hyperglycaemia (neuropathy,
nephropathy, retinopathy and cateracts) rather than the control of hyper-
glycaemia per se. They were envisaged as synergistic additions to standard
diabetic treatments and were a group of structurally different compounds
that bound to aldose reductase, thus inhibiting the conversion of glucose to
sorbitol. This inhibition circumvents the build-up of sorbitol in the cellular
milieu and the resulting decrease in myoinositol uptake. It is this decreased
myoinositol uptake in nerve tissue that appears to be responsible for a variety
of metabolic and electrophysiologic alterations including alterations in nerve
conduction (BOULTON et al. 1990; GREEN and JASPAN 1990; KIRCHAIN and
RENDELL 1990; MASSON and BOULTON 1990; ZENON et al. 1990; VAN GERVEN et
al. 1992).
Tolrestat is the only one of the original ARls still undergoing clinical trial.
Results have been encouraging, but by no means definitive, for some aspects
of diabetic neuropathy. Improvements have occurred in paraesthesia and
neuropathy, but unfortunately not in pain symptoms. Ongoing studies con-
tinue (see review TSAI and BURNAKIS 1993).
A more recent and possibly synergistic treatment with established oral
hypoglycaemic agents is acarbose, an a-glucosidase inhibitor. This is a new
type of antidiabetic agent, which because of its similarity in structure to dietary
disaccharides inhibits the enzymes responsible for the digestion of carbohy-
drate. The effect is to delay the digestion of starch and sucrose. In diabetic
patients this results in a lowering of postprandial hyperglycaemia and smooth-
Synergistic Drug Interactions 245

ing of daily blood glucose fluctuations. The antidiabetic effect does not involve
the stimulation of insulin secretion. Gastrointestinal problems are the most
commonly reported adverse effects in clinical trials. These include flatulence,
abdominal distension and diarrhoea. They are caused by fermentation of
unabsorbed carbohydrate in the bowel (CLISSOLD and EDWARDS 1988; NEW
MEDICINES 1994).
Currently, acarbose is licensed in the United Kingdom solely for the
treatment of non-insulin-dependent diabetes inadequately controlled by diet
alone or diet and oral hypoglycaemic agents. Although it has well-documented
antidiabetic properties, its precise place in therapy remains to be established.
It is not clear which patients should receive acarbose in place of, or in addition
to, established oral hypoglycaemic agents. It has been suggested that acarbose
may be useful for patients receiving an oral hypoglycaemic agent at the top of
the dosage range as an alternative strategy to commencing insulin. Further
clinical experience is required to establish whether it should be used in place
of traditional oral agents or whether it will act synergistically in combination
with them.

J. Cancer Chemotherapy
It has become standard practice to treat various malignant diseases with
cocktails of intensive combination regimens. For example in the treatment
of myeloma a combination of melphalan, cyclophosphamide, carmustine,
vincristine and prednisolone (M2) has been used. Other regimens combining
vincristine, doxorubicin and a corticosteroid, either dexamethasone (VAD) or
methylprednisolone (VAMP) have been developed.
In Ewing's sarcoma a protocol combining chemotherapy with cyclophos-
phamide, vincristine, dactinomycin and doxorubicin (VACA) with local radio-
therapy has achieved a 5-year survival rate. Other combination regimens have
been used for special tumours, e.g., CHAP 5, cisplatin, cyclophosphamide,
altretamine and doxorubicin combined with surgery for ovarian carcinoma.
Two of the better known cocktails are the MOPP regimen of mustine,
vincristine, procarbazine and prednisolone used in Hodgkin's disease, and the
COAP combination of cyclophosphamide, vincristine, ara-AC (fazarabine)
and prednisolone used in acute leukaemias.
Cocktails for use in malignant conditions have been developed by special-
ists in the field, partly on good therapeutic experience and partly empirically.
Where there is synergy, the mechanisms are often obscure. In many cases the
regimens have been developed from the philosophy of throwing into the
treatment everything in the pharmacopoeia that logically might work, and
avoiding two agents with the same mechanism of action on the malignancy.
From that initial experimental philosophy over the last 20 years there has
developed a rational approach to the evolution of the chemotherapeutic
cocktails.
246 J.P. GRIFFIN and P.F. D'ARCY

K. Beneficial Interactions: A Philosophical Approach

The examples of beneficial drug-drug interactions that have been described
in the previous sections of this chapter emphasise that synergistic drug
interactions when they occur can be of clinical value in drug therapy. Apart
from these major interactions that have already been cited, there are many
others that have the potential for clinical advantage. Food-drug interactions
can be beneficial in some instances, for example, taking propranolol or
metoprolol with food improves their bioavailability (MELANDER et al. 1977)
and the absorption of cefuroxime axe til is improved in the presence of food
(RIDGWAY et al. 1991; REYNOLDS 1993). Improvements are also seen in the
absorption of the antileprosy agent clofazimine (HOLDINESS 1989; REYNOLDS
1993), the urinary antiseptic nitrofurantion (ROSENBERG and BATES 1976),
the antifungal itraconazole (V AN PEER et al. 1989) and the antifungal
griseofulvin in the presence of fatty foods (CROUNSE 1961) or milk (GINSBURG
et al. 1983).
There are other examples where drug-drug interactions can improve the
efficacy of therapy, for example, the treatment of erythropoietin resistance
with cyclosporin (ALMOND et al. 1994). Cimetidine binds to microsomal P450
enzymes and inhibits the oxidative phase of hepatic drug metabolism, thus
potentiating the effect and/or duration of a variety of drugs. These drugs
include anticoagulants (e.g., warfarin), theophylline and aminophylline, some
benzodiazepines, anti-arrhythmic agents (e.g., lignocaine and quinidine),
anticonvulsants (e.g., carbamazepine and phenytoin), some beta-blockers,
narcotic analgesics, and tricyclic antidepressants (e.g., nortriptyline).
Cimetidine also increases the bioavailability of fluorouracil by over 70% with-
out evidence of increased toxicity (see GRIFFIN et al. 1988).
It is understandable to think that increasing the plasma concentrations of
a primary drug, for example by cimetidine, will automatically and conse-
quently increase the probability of increased toxicity of that drug. However,
this may not always occur and if it does not then the dose-sparing effect on the
primary drug would have both a clinical and cost advantage. There could
therefore be much advantage in searching for a chemical substance which
would reversibly inhibit the P450 oxidase systems without itself having a
distinct pharmacological or toxic action. However, examination of this possi-
bility lies in the future.
Today, in practical terms, the "eldorado" of synergy is very often, but
fortunately not always, as elusive as it was at the time of Mithridates. A
healthy trend towards scepticism with a touch of optimism probably
summarises our current viewpoint.

References
Alberti KGMM, Hockaday TDR (1987) Diabetes mellitus. In: Weatherall DJ,
Ledingham JGG, Warrell DA (eds) Oxford textbook of medicine vol 1, 2nd edn.
Oxford Medical, Oxford, pp 9.51-9.101
Synergistic Drug Interactions 247

Almond MK, Tailor D, Kelsey SM, Cunningham J (1994) Treatment of erythropoietin

resistance with cyclosporin. Lancet 343:916-917
Boulton AJM, Levin S, Comstock J (1990) A multicenter trial of the aldosereductase
inhibitor, tolrestat, in patients with symptomatic diabetic neuropathy.
Diabetologia 33:431-437
British National Formulary (1981) No 1. British Medical Association and Royal
Pharmaceutical Society of Great Britain, London
British National Formulary (1994) No 27. British Medical Association and Royal
Pharmaceutical Society of Great Britain, London
Chadwick D (1993) Seizures, epilepsy and other episodic disorder. In: Walton J
(ed) Brain's diseases of the nervous system, 10th edn. Oxford Medical, Oxford, p
718
Clissold SP, Edwards C (1988) Acarbose: a preliminary review of its pharmacodynamic
and pharmacokinetic properties, and therapeutic potential. Drugs 35:214-243
Crounse RG (1961) Human pharmacology of griseofulvin: the effect of fat intake on
gastrointestinal absorption. J Invest Dermatol 37:529-533
Ginsburg CM, McCracken GH Jr, Petruska M, Olsen K (1983) Effect of feeding on
bioavailability of griseofulvin in children. J Pediatr 102:309-311
Green A. Jaspan J (1990) Treatment of diabetic neuropathy with inhibitors of the
aldose reductase enzyme. J Diabetic Complications 4:138-144
Griffin JP (1994) Mithridates VI of Pontus, the first experimental toxicologist. Adv
Drug React Toxicol Rev 14:1-6
Griffin JR, Wyles M (1991) Epilepsy, towards tomorrow. Office of Health Economics,
London
Griffin JP, D' Arcy PF, Speirs CJ (1988) A manual of adverse drug interactions, 4th edn.
Wright, (Bulterworth) London
Gustavson LE, Kaiser JF, Mukherjee DX, Schneck DW (1994) Evaluation of the
pharmacokinetic drug interactions between clarithromycin and omeprazole
(Abstr). 10th World Congress of Gastroenterology, Oct 2-7, Los Angeles
Heberden W (1745) Antitherica, essay on Mithridatium and Theriac. London
Hitchings GH, Elion GB, Vanderwerff H, Falco EA (1948) Pyrimidine derivatives as
antagonists of P.G.A. J Bioi Chern 174:765-766
Holdiness MR (1989) Clinical pharmacokinetics of clofazimine: a review. Clin
Pharmacokinet 16:74-85
Hoskings SW, Ling TKW, Chung S, Yung MY, Cheng AFB (1994) Duodenal ulcer
healing by eradication of Helicobacter pylori without anti acid treatment. Lancet
343:508-510
Kirchain W, Rendell M (1990) Aldose reductase inhibitors. Pharmacotherapy 10:326-
336
Knatterud GL, Meinert CL, Klimt C, Osborne RK, Martin DB (1971) Effects of
hypoglycemic agents on vascular complications in patients with adult-onset
diabetes. IV. A preliminary report on phenformin results. JAMA 217:777-784
Lasseter KC, Levey GS, Palmer RF, McCarthy JS (1972) The effects of sulfonylurea
drugs on rabbit myocardial contractility, canine Purkinje fiber automaticity, and
adenylcyclase activity from rabbit and human hearts. J Clin Invest 5:2429-2432
Logan RPH (1994) Clarithromycin may prevent ulcer recurrence. 7th Workshop on
Gastroduodenal Pathology and Helicobacter pylori, Sept 30, Houston
Masson EA, Boulton AJM (1990) Aldose reductase inhibitors in the treatment of
diabetic neuropathy: a review of the rationale and clinical evidence. Drugs 39:190-
202
Medicines Resource Centre (1994) Combination diuretics. MeRec Bull 5(10)
Melander A, Danielson K, Schersten B, Wah lin E (1977) Enhancement of the
bioavailability of propranolol and metoprolol by food. Clin Pharmacol Ther
22:108-112
New Medicines (1994) Acarbose - an a-glucosidase inhibitor. Int Pharm J 8:11-12
Prout TE (1975) A progress report on the University Group Diabetes Program. Int J
Clin PharmacoI12:244-251
248 J.P. GRIFFIN and P.F. D'ARCY: Synergistic Drug Interactions

Reynolds JEF (ed) (1993) Martindale, the extra pharmacopoeia. Pharmaceutical Press,
London, pp 137-138, 151-152
Ridgway E, Stewart K, Rai G, Kelsey MC, Bielawska C (1991) The pharmacokinetics
of cefuroxim axetil in the sick elderly patient. J Antimicrob Chemother 27:663-668
Rosenberg HA, Bates TR (1976) The influence of food on nitrofurantoin bio-
availability. Clin Pharmacol Ther 20:227-232
Sales JEL, Sutcliffe M, O'Grady F (1972) Cephalexin levels in human bile in presence
of biliary tract disease. Br Med J 3:441-443
Shanks RG (1979) Cardiac dysfunction. In: D'Arcy PF, Griffin JP (eds) Iatrogenic
diseases, 2nd edn, Oxford Medical, Oxford, pp 116-131
Sneader W (1985) Drug discovery: the evolution of modern medicines. Wiley,
Chichester
Soler NG, Pentecost BL, Bennett MA, Fitzgerald MG, Lamb P, Malins JM (1974)
Coronary care for myocardial infarction in diabetes. Lancet 1:457-477
Tsai SC, Burnakis TG (1993) Aldose reductase inhibitors: an update. Ann
Pharmacother 27:751-754
University Group Diabetes Program (1970) A study of the effects of hyperglycemic
agents on vascular complications in patients with adult onset diabetes. II.
Mortality results. Diabetes 19 [Suppl]:789--830
Van Gerven J, Lemkes H, van Dijk J (1992) Long-term effects of tolrestat in
symptomatic diabetic sensory polyneuropathy. J Diabetes Complications 6:45-48
Van Peer A, Westenborghs R, Heykants J, Gasparini R, Gauwenbergh G (1989) The
effect of food and dose on the oral systemic availability of itraconazole in healthy
subjects. Eur J Clin Pharmacol 36:423-426
Zenon G, Abobo C, Carter B, Ball D (1990) Potential use of aldose reductase
inhibitors to prevent diabetic complications. Clin Pharm 9:446--457
CHAPTER 9
In Vitro Drug Interactions
J.e. McELNAY and e.M. HUGHES

A. Introduction
In vitro drug interactions may be defined as those interactions which occur
outside the body. Drug interactions which will be discussed in this chapter are
those which occur between drugs due to reasons of incompatibility (e.g., drug-
drug interactions in an intravenous infusion), due to interaction of a drug with
its packaging (e.g., drug binding to an infusion bag), due to loss of drugs during
laboratory analyses (e.g., binding to laboratory equipment) or due to changes
in the bioavailability of drugs when the formulation is altered (McELNAY and
D'ARCY 1980).

B. Incompatibility Interactions
This form of interaction is often described as an incompatibility between two
agents and may be physical or chemical in nature. The interactions often occur
in solution, e.g., following addition of drugs to intravenous fluid containers,
after mixing drugs in syringes or in liquid preparations for oral or topical
administration. Physical incompatibility may be due to immiscibility or insolu-
bility. For example, addition of high concentrations of electrolytes to mixtures
in which the vehicle is a saturated aqueous solution of a volatile oil causes the
oil to separate and collect as a surface layer. This is the case with potassium
citrate mixture BPC, in which the large quantity of soluble solids salts out the
lemon oil; quillaia tincture is therefore added as a suspending and emulsifying
agent to prevent this (CARTER 1975).
Chemical incompatibility may be caused by pH change, decomposition or
complex formation. A prime example of this type of interaction involves the
barbiturates, e.g., pentobarbitone and phenobarbitone. Solutions of the salts
of these compounds are very alkaline and incompatible with acids, acidic salts
and acidic syrups, e.g., lemon syrup, which may precipitate the corresponding
insoluble barbituric acid derivative (CARTER 1975). The BRITISH PHARMACEUTI-
CAL CODEX (LUND 1994) documents a number of physical and chemical inter-
actions between drugs and other medicinal products; these are summarised in
Table l.
Many such incompatibilities are no longer viewed as being particularly
problematic as far as oral formulations are concerned since easy solutions to
250 J.e. McELNAY and e.M. HUGHES

Table 1. Examples of in vitro drug-drug incompatibilities from the British

Pharmaceutical Codex (LUND 1994)

Mechanism of interaction Example

Precipitation of drug on Cosolvents may be used to increase the solubility of

dilution of solutions drugs which are poorly aqueous-soluble, e.g., digoxin
which contain cosolvents and diazepam. These cosolvents may, however, reduce
the solubility of other constituents such as gums,
polymers and sugars. Dilution may result in
precipitation of drug unless the extent of dilution is
such that the final concentration is below its solubility in
water.
Precipitation by The solubility of an organic compound or excipient may
salting-out be reduced by addition of a salt. Sodium chloride and
or potassium chloride decrease the solubility of benzoic
acid; reconstitution of erythromycin lactobionate with
saline rather than water may result in precipitation or
gel formation. Such phenomena may be attributed to
the effects of the ionic size and valency of salts on the
structure of water or to competition between salts and
organic solutes for water molecules.
Ionic interactions Examples of cationic-anionic interactions resulting in
possible precipitation from solution are promethazine
hydrochloride-thiopentone sodium, kanamycin sulphate-
sulphadiazine sodium.
Formation of complexes Tetracycline derivatives form stable complexes with
metal ions, e.g. calcium, magnesium and ferric iron,
resulting in reduced absorption of tetracycline.
Adsorption on solid Antidepressants, antihistamines, f3-blockers and
particles benzodiazepines may be adsorbed in vitro on clays such
as kaolin or suspended antacids. Such interactions may
arise due to physical adsorption (drug is bound by weak
Van der Waal forces) or chemical adsorption (drug is
bound by strong valence forces).

the problems can usually be found. There are often, however, many contem-
porary reports in the pharmaceutical literature on drug interactions due to
incompatibility in intravenous fluids. This can either be due to an interaction
between two or more drugs added to intravenous fluids prior to administration
to patients, e.g., in the case of cancer chemotherapy or due to a single added
drug being incompatible with the intravenous vehicle. GRIFFIN and D' ARCY
(1988) have prepared a list of such incompatibilities, for example, ampicillin
should not be mixed with amino acids [constituents of total parental nutrition
(TPN) regimens] as immunogenic and allergic conjugates could be formed;
amphotericin is incompatible with all electrolytes and, thus, must be reconsti-
tuted with water and infused with 5% glucose (GRIFFIN and D'ARCY 1988).
ALLWOOD (1994) described the problem of drug-drug co-precipitation, which
may be particularly problematic in intravenous therapy. Such interactions
In Vitro Drug Interactions 251

arise due to the mixing of organic anions and cations, which form ion pairs.
Examples of drugs at risk of forming ion pairs with other drugs include
gentamicin and other aminoglycosides, with heparin and some cephalosporins.
One of the areas of medicine where simultaneous co-administration of a
number of intravenous medicines is a clinical necessity is cancer chemo-
therapy. FOURNIER et al. (1992) studied the effect of a commercial formulation
of 5-fluorouracil on the stability of cisplatin and carboplatin to determine
whether the drugs could be mixed in containers or intravenous lines. When
cisplatin was incubated with 5-fluorouracil, high-performance liquid chroma-
tography (HPLC) studies demonstrated a rapid disappearance of the parent
platinum compound, the extent of degradation being 75% after 3.5h. It was
found that the degradation was in fact caused by an interaction between
cisplatin and trometamol, the excipient used in the 5-fluorouracil formulations
in some countries to buffer the solution at pH 8.2. This interaction resulted in
complete inhibition of the antitumour activity of cisplatin in mice. When
cisplatin was incubated at the same pH in trometamol-free sodium hydroxide
solutions, the parent compound was transformed into an active compound that
was toxic to mice. Degradation of carboplatin-trometamol admixtures was
similar to that found for cisplatin, but occurred at a slower rate. As in the case
of cisplatin, incubation of carboplatin in a sodium hydroxide solution resulted
in toxic effects. It was concluded that both carboplatin and cisplatin were
incompatible with 5-fluorouracil formulations containing trometamol.
Due to the possibility of incompatibility with constituents, as a general
rule no drugs should be added to TPN fluids. To avoid direct contact between
TPN fluids and drugs administered by the intravenous route, COLLINS and
LUTZ (1991) have examined the co-administration of phenytoin and TPN
mixtures in multilumen catheters. These devices can be used to reduce the
need for frequent venipuncture and provide a method of administering incom-
patible drugs. The study utilised an in vitro model flow system to examine the
physicochemical phenomena that occurred when these two incompatible enti-
ties (phenytoin and TPN fluid) were simultaneously administered through
multilumen catheter systems into a circulating fluid (sodium chloride, sodium
bicarbonate and albumin in solution). Double- and triple-lumen catheters
were used as shown in Fig. 1. Video recordings were made of the drug flow and
assays of phenytoin concentration were performed on samples of the circulat-
ing fluid. White clouds of phenytoin precipitation were seen near the tip of the
double-lumen catheter, but not the triple-lumen device. Infusion through the
former resulted in an average of 6% loss of phenytoin which, on microscopic
examination, appeared as spindle-shaped crystals. In some cases, millimetre-
size fragments of phenytoin precipitate were seen to dislodge from the tip of
the double-lumen catheter. It was suggested that the interaction was due to the
close proximity of the two orifices at the end hole of the double-lumen cath-
eter, which permitted mixing of the two effusing streams of the incompatible
agents. The staggered orifices of the triple-lumen catheter reduced this inter-
action greatly; no crystal fragments were observed, although a thin coating of
252 J.C. McELNAY and C.M. HUGHES

A.
Double
lumen
Call'leter

B
Tnpl~ Lumen
Catheter

Fig. 1. Drawings of tip of two types of catheters used in study of concomitant admin-
istration of phenytoin and TPN fluid. (After COLLINS and LUTZ 1991)

white film did form near the opening of the proximal and middle side holes of
the catheter. Although the authors recognised the value of these multilumen
devices, they did advise caution in their use, particularly in the clinical setting,
with combinations of incompatible drugs. To be forewarned is to be forearmed
and indeed the best approach to prevention of drug-drug incompatibility
interactions is to use published information on the physical and chemical
properties of medicinal agents which are to be used concomitantly and avoid
their coming into direct contact.

C. In Vitro Drug Interactions with Pharmaceutical

Packaging and Intravenous Administration Equipment
One of the criteria which has been designated as essential in relation to
pharmaceutical packaging is that it should not interact with a drug product
and, conversely, that the drug product should not interact with the packaging.
Careful consideration should therefore be given to the materials used in
pharmaceutical packaging. Incompatibilities between drugs and such materi-
als have been well documented in the literature. The two materials which are
most widely used in packaging, particularly in intravenous therapy, are plastics
and, to a lesser extent, glass.

I. General Properties of Plastics

Plastics are defined as a wide range of solid composite materials which are
largely organic, usually based upon synthetic resins or upon modified polymers
In Vitro Drug Interactions 253

of natural origin and possessing mechanical strength. The physical, chemical

and mechanical properties of a plastic material are determined by its chemical
structure, molecular weight range and alignment of the resin, and the type and
concentration of additives. Selection of the monomer, for example, ethylene,
propylene, glucose, vinyl chloride, determines the chemical structure and the
type of substituents in and/or on the polymer chain (AUTAIN 1963; COOPER
1974; BIRELY and SCOTT 1982; BRYDSON 1982). Table 2 lists a number of

Table 2. Polymer compounds which are widely used in plastic pharmaceutical

packaging and intravenous administration equipment. (After YAHYA et al. 1986)

Chemical name Plastic type Repeating unit of the polymeric

structure of the plastic

Cellulose acetate Amorphous thermoplastic

Cellulose propionate Amorphous thermoplastic CH2 -OR

I
CH-O
-CH/ ""CH-O-
""yH-yH/
RO OR
R= -CO-C2HS

Ethylvinyl acetate Amorphous thermoplastic -CH2-CH-

6-CO-CH2- CH 3

Methacrylate Amorphous thermoplastic Co-polymer

butadiene styrene

Polyethylene Partially crystalline

thermoplastic

Polypropylene Partially crystalline -CH -CH-

thermoplastic 2 I
CH3

Polystyrene Crystalline thermoplastic

Polyvinylchloride Amorphous thermoplastic -CH -CH-

2 I
Cl
254 J.e. McELNAY and e.M. HUGHES

polymer compounds which are widely used in plastic pharmaceutical packag-

ing and intravenous administration equipment.

II. General Properties of Glass

Glass containers are particularly useful for liquid preparations owing to their
rigidity, their superior protective qualities and their ability to allow easy in-
spection of the contents. Glass is impermeable to air and moisture, inert to
most medicinal products and can be coloured to protect the contents from
light of certain wavelengths (LUND 1994).
For most drug storage purposes, ordinary so-called soda glass or white
flint glass containers are used. Amber glass is used for light-sensitive drugs and
its composition differs little from white flint glass, except that a small propor-
tion of iron oxide is added under strongly reducing conditions. Although glass
is more resistant to chemical attack than other packaging materials, it is not
entirely inert and, when it is in prolonged contact with water, alkali tends to be
extracted. Alkaline solutions attack glass more rapidly than water and the
higher the alkali content of the glass the more rapid is this effect. Thus for
preparations which are liable to attack glass, a glass of low alkali content must
be used.
Glass completely devoid of alkali is not practical but in borosilicate glass
such as Pyrex the soda content is minimal at about 3.5%, the silica content is
around 80% and the boric acid content is approximately 13%. Such glass is
neutral and has very high chemical and thermal resistance. It is, however, hard
and more difficult to process than glass with a higher alkali content. Borosili-
cate glass is widely used in laboratories where inert glass is required.

III. Mechanisms of Interaction with Pharmaceutical Packaging

There are a number of mechanisms which are considered central to any
interpretation of drug-packaging interactions. These are:
1. Sorption - a term which describes both adsorption of a drug to a surface
and the absorption of a drug into the matrix of the material under
examination.
2. Leaching - this is the process by which components of the packaging
material migrate into the medicine.
3. Permeation - this process involves the transfer of drug through the packag-
ing material to the external surface. This can be followed by evaporation of
the drug.
4. Polymer modification - in some cases, drugs can interact with the polymer
in such a way as to modify the chemical structure of the polymer which may
lead to changes in its physical properties.
Each mechanism will be discussed and illustrated with an appropriate ex-
ample. Table 3 summarises a number of documenter drug-packaging interac-
 ......
:::
$
;::;-
o
tl
8
{JQ

Table 3. Examples of drug-packaging interactions ......

:::....
(l)

Drug Example of interaction Management of interaction Reference Il>

(')
"'
C.
o
Antineoplastic agents :::
en
Carmustine Sorbed to PVC Use glass containers BENVENUTO et a1.
Bleomycin Sorbed to PVC Use glass containers (1981)
Mitomycin Sorbed to PVC and glass
Doxorubicin Sorbed to glass
Fluorouracil Sorbed to glass
Methotrexate Sorbed to glass when prepared Use plastic containers and protect CHEN and CHIOU
in methanol or ethanol from light (1982)
Vinblastine Sorbed to PVC and cellulose Use alternative plastic materials, e.g., McELNAY et a1.
propionate methacrylate butadiene styrene (1988)
Immunosuppressant agents
FK 506 Sorbed to PVC and leaching Use polyolefin or glass containers TAORMINA et a1.
of plasticiser from PVC (1992)
Hypnotic and psychotropic
drugs
Diazepam Sorbed to PVC Use glass or alternative plastic materials, e.g., PARKER and
polyethylene and polypropylene McCARA (1980)

N
Ul
Ul
 N
Ul
0\

Table 3. Continued

Drug Example of interaction Management of interaction Reference

Chlopromazine Sorbed to PVC Use alternative form of plastic for AIRAUDO et al.
Clomipramine Sorbed to PVC administration, e.g., Stedim infusion bags (1993)
Chlorazepate dipotassium Sorbed to PVC with polyethylene lining
salt
Vitamins
Vitamin A Sorbed to PVC Use vitamin A palmitate which CHIOU and
does not sorb to PVC MOORHATCH
(1973)
Vitamin E Sorbed to plastic materials in Use glass DROIT et al. (1991)
infusion sets (material not
specified)
Vitamin D (ca1citriol) Sorbed to PVC Increase amount of surfactant used in LI et al. (1992)
formulation as this reduces amount of free
drug available for sorption. Avoid PVC ......
Miscellaneous agents (1
Miconazole Sorbed to PVC Use glass; if PVC must be used, drug must HOLMES and
be prepared and administered immediately ALDOUS (1991) ~
t!j
Bupivacaine Sorbed to glass and plastic Avoid alkalinisation as sorption effects BOMHOMME et al. t"'

after alkalinistation conteract any clinical benefits in adjusting (1992) ~

(prolongs release of pH ~
t:l
bupivacaine and thus 0-
increases duration of action) n
Warfarin Sorbed to PVC Use glass KowALUK et al. ~
(1981)
::r:
c::
Cl
:Ii
ttl
CfJ
In Vitro Drug Interactions 257

tions, the underlying mechanism of action and potential solutions to overcom-

ing these problems.

1. Sorption
Sorption is probably the most important underlying mechanism responsible
for a number of drug interactions associated with pharmaceutical packaging.
This term describes loss of drug to the interacting material and includes
adsorption to the surface plus absorption of the drug into the body of the
material. Adsorption is characterised by rapid and substantial initial drug loss
associated with binding to the surface, leading to surface saturation. Adsorp-
tion can occur to all solid surfaces irrespective of the nature of the material
(ALLwooD 1994). Absorption is defined as the migration of the drug into the
matrix of the interacting material (usually plastic) and an equilibrium becomes
established between liquid and solid phases. The process develops more slowly
than adsorption, but saturation occurs eventually, although it may take many
hours (ALLwooD 1994). The plastic that is most often implicated in such
interactions is polyvinyl chloride (PVC), although other materials can also be
involved.
When examining sorption interactions, the physicochemical properties of
the drug in question must be considered: the extent of ionisation and lipid
solubility have been proposed as two determining factors of sorption by plastic
infusion bags (KOWALUK et al. 1981). A number of models have been devel-
oped to predict solute sorption by PVc. ROBERTS et al. (1991) estimated the
time course of the sorption of drugs using a diffusional model in which the
plastic was assumed to act as an infinite sink. This model appeared to be
suitable for estimation of storage times relevant to clinical usage and enabled
the magnitude of the uptake in a specified time to be described by a single
parameter known as the sorption number. This sorption number was defined
by the plastic-infusion solution partition coefficient, the diffusion coefficient in
the plastic, the fraction of drug unionised in solution, the volume of the
infusion solution and the exposed surface area of the plastic. This parameter
(sorption number) could be used to predict the effects of a number of factors,
e.g., time, plastic surface area, solution volume and solution pH on fractional
drug loss. JENKE (1993) developed a similar model in which storage time
proved to be a critical factor in predicting sorption; at longer storage times, it
was found that the infinite sink approach as described by ROBERTS et al. (1991)
was inaccurate, while at shorter storage times the sorption profile for a solute
can be estimated from its hexane-water and octanol-water partition coeffi-
cients via a simple diffusion-based expression. It was noted that differences
between observed and predicted behaviour illustrated the problems in a theo-
retical approach; thus, it is only in the experimental or clinical environment
that accurate assessment relating to drug sorption can be made.
Not all sorption interactions will be clinically significant. The schematic
representation of an arbitrary time scale (Fig. 2) illustrates that those inter-
 258 J.e. McELNAY and C.M. HUGHES

•+
Nitroglycerin Vitamin A Benzodiazepine Lignocaine

Itt Chlormethiazole
t
24 48
Insulin Diazepam Warfarin

Fig. 2. Time-scale for clinically significant losses of drug due to sorption to plastics
materials (ALLWOOD 1995, personal communication)

actions occurring close to time 0 are clinically significant, whereas those occur-
ring at the other end of the scale may be considered insignificant. Insulin and
the nitrates are examples of drugs which are sorbed to a clinically significant
degree; interactions involving these drugs are discussed below.

a) Insulin
The non-specific surface binding of insulin from dilute solutions was first
reported by FERREBEE et al. (1951), who showed that the polypeptide was
strongly adsorbed by laboratory glassware. Since then, further work has shown
that insulin also binds to siliconised glassware, paper and polypropylene
(NEWERLY and BERSON 1957; HILL 1959; PETIY and CUNNINGHAM 1974).
WIDEROE et al. (1983) evaluated the degree of insulin adsorption to plastic
containers during continuous ambulatory peritoneal dialysis (CAPD). Ap-
proximately 65 % of insulin which was added to the dialysis fluid was sorbed to
the plastic container.
A series of studies by TWARDOWSKI et al. (1983a-c) examined this sorption
phenomenon in greater depth. Binding of insulin to the plastic of dialysis
solution containers was instantaneous and apparently not influenced by time
of contact. Further experimental work revealed that insulin binding could be
partly reversed by simple washing with water (TWARDOWSKI et al. 1983b). A
third study demonstrated that insulin sorption was influenced by temperature:
more insulin was bound at 37°C than at 24°C (TWARDOWSKI et al. 1983c).
McELNAY et al. (1987) studied the binding of human insulin to three types
of burette intravenous administration sets - a Standard set (cellulose propi-
onate burette and PVC tubing), an Amberset (cellulose propionate burette
and PVC tubing) and a Sure set (methacrylate butadiene styrene burette and
polybutadiene tubing) over a 6-h period. The latter administration set had
been previously shown to resist sorption of a number of drugs (LEE 1986).
However, in this case, the Sureset bound insulin more extensively than the
Standard set or Amberset as illustrated in Fig. 3. It was suggested that, in the
clinical setting, this adsorption would present a particular problem since dur-
ing intravenous infusion the concentration of drug is relatively low and the
surface area available for binding is relatively high. This non-specific adsorp-
tion of insulin can best be approached by administering insulin in a small
volume via a syringe pump as the surface area is much reduced compared with
the total amount of insulin present or, alternatively, more drug than is re-
quired is added to the infusion container and blood sugar levels closely moni-
In Vitro Drug Interactions 259

25
20
'"
E
0 15
0

-- 10
I/)

::J
~ 5
0
o 0.083 0.5 2 4 6

a Time (Hours)

c: 25

-cE
.2
..,-
.. '" 20
Gl o 15
go
01/)
0--
10
c: =!
3;':; 5
.='" 0 ~ ~ r1r.;:l
~
o 0.083 0.5 2 4 6

b Time (Hours)

Fig. 3. a Concentration versus time profile for human insulin when stored in burette
chambers of a standard set (0), Amberset (~) and Sureset (~ burette administration
sets. • represents mean control data at zero time. (Maximum coefficient of variation
within replicate data points was 3.2%.) b Concentration (± SD) versus time profile for
human insulin when stored in administration tubing from a standard set (0), Amberset
(~) and Sure set (~) burette administration sets. • represents mean control data at
zero time. (After McELNAY et al. 1987)

tored. It has also been suggested that human serum albumin or hydrolysed
gelatin (polygeline) can be used as a carrier for insulin, since the binding to
these agents exceeds that of the binding to surfaces (SCHILDT et al. 1978). This
approach is, however, not used routinely.

b) Nitrates
A number of studies have reported interaction of a range of nitrates, e.g.,
glyceryl trinitrate and isosorbide dinitrate with packaging and intravenous
delivery equpiment. A number of mechanisms of interaction appear to take
place simultaneously with the nitrates including adsorption, absorption and
permeation.

c) Glyceryl Trinitrate (Nitroglycerin)

The physical instability of glyceryl trinitrate can be partly attributed to sorp-
tion of drug from solution to plastic containers, YUEN et al. (1979) published
the results of a study in which the equilibrium and kinetic techniques were
260 J.e. McELNAY and e.M. HUGHES

used to examine the loss of drug efficacy. It was concluded that glyceryl
trinitrate was removed from aqueous solution by plastic via an absorptive
process, suggesting a degree of penetration into the plastic. It is also likely, due
to the volatile nature of glyceryl trinitrate, that it penetrates through the
plastic and is lost by evaporation from the plastic air interface.
The loss of glyceryl trinitrate was confirmed by BAASKE et al. (1980), who
studied the effect of intravenous filters, containers and administration sets on
glyceryl trinitrate efficacy. Filters decreased glyceryl trinitrate concentrations
by 2%-55%, whereas storage in glass bottles over 48h had no effect. Storage
in plastic intravenous infusion bags led to substantial concentration decreases
that were related to surface contact area and temperature. Glyceryl trinitrate
solution was also infused through an administration set, resulting in an imme-
diate and substantial reduction in drug concentration. Using a theoretical
approach, MALICK et al. (1981) described a model of glyceryl trinitrate loss in
which the drug is adsorbed onto the material surface and then partitioned into
the plastic.
Work has continued on the glyceryl trinitrate sorptive phenomenon.
LOUCAS et al. (1990), for example, examined the sorptive properties of glyceryl
trinitrate in relation to PVC, with particular reference to the role of the
admixture vehicle. Initially (first 10min of sampling) the loss of glyceryl
trinitrate to the PVC administration set was greatest from dextrose admix-
tures, intermediate for water admixtures and least for saline admixtures. Be-
tween 15 and 20 min of the experimental period, the greatest loss of glyceryl
trinitrate was noted for saline. It was concluded that glyceryl trinitrate avail-
ability in admixtures in contact with PVC was dependent on the ionic strength
of the vehicle and the time at which measurements were made.
The loss of glyceryl trinitrate during passage through a PVC or a polyeth-
ylene infusion set has also been investigated in relation to concentration and
flow rate (HANSEN and SPILLUM 1991). The concentration of glyceryl trinitrate
had no influence on the loss of drug, although the greatest loss occurred at the
slowest rate of infusion. A loss of 70% of glyceryl trinitrate during infusion
using PVC apparatus was found, whereas the loss when using the polyethylene
set did not exceed 15% over 8h.
As PVC appears totally unsuitable as a material in which to store glyceryl
trinitrate, other forms of plastics have been used. SALOMIES et al. (1994)
measured the sorption of glyceryl trinitrate, diazepam and warfarin sodium (in
normal saline) by a new polypropylene-lined infusion bag (Softbag) and com-
pared this to sorption in glass bottles and PVC bags. The containers were
stored at room temperature without protection from light for 24h (diazepam)
and 120h (glyceryl trinitrate and warfarin). The three drugs did not demon-
strate any sorption to glass bottles or the Softbag containers although sorption
was high to PVC (Fig. 4). The polypropylene-lined Softbag therefore provides
a useful approach in the avoidance of sorption of these agents.
ALTAVELA et al. (1994) viewed that loss of glyceryl trinitrate onto admin-
istration sets may have little impact on the care of patients as infusions of the
In Vitro Drug Interactions 261

100

80
roo-----------I
e,
riia

.".,,~
o
0.

o
'"
.... 60
o
~ 40

20
o {L_ _ _ _- L____- L____- L____- L____ __ ~ ~

o 5 10 15 20 25
a Time (hours)

100 ift3:..--~-g-~--O-g

c
.;: 80 ~-e
'"
U
\
e

""
>.
60

-
CI
~
z 40 e'-...........
e_e_
~ e-e
20

o ~ __ ~ __ ~ ____ ~ __ _ L_ _~_ _ _ _~~

o 20 40 60 80 100 120
b Time (hours)

- - 0 - - - 0 ___ 0 ___ 0 ___ 0

100 -"7--"7--"-"7---'11

--"
E T"
~

"0
0
80 e - T _ " ___ T
en \
e

"" --e---. --e_.

...0
C
60
...c e
:;t
40
~

20
o{
0 20 40 60 80 100 120
C Ti m e (h 0 u rs)

Fig. 4. 3 Concentration of diazepam (measured by UV spectrophotometry) as a func-

tion of time in glass bottles (0) and in PVC (e) and Softbag (\7) infusion bags. Values
expressed as means of six determinations. Coefficient of variation (CV) values varied
between 0.50% and 3.26%. b Concentration of nitroglycerin (measured by HPLC) as
a function of time. Values expressed as means of six determinations. CV values varied
between 0.71 % and 2.21 %. Symbols as in 3. c Concentrations of warfarin sodium
(measured by UV spectrophotometry) as a function of time in glass bottles (0) and in
PVC pH 4.9 (e), PVC pH 5.6 (~) and Softbag (\7) infusion. Values expressed as means
of two (pH 5.6) or four (pH 4.9) determinations. CV values of the latter varied between
0.38% and 3.09%. (After SALOMIES et al. 1994)
262 J.e. McELNAY and e.M. HUGHES

drug are adjusted for the patient's response and thus drug loss may be of little
significance. Another consideration was the expense of polyethylene sets com-
pared to PVC sets; thus this group randomly assigned patients with ischaemic
heart disease to receive glyceryl trinitrate solution through either a polyethyl-
ene administration set or a PVC set. It was found that patients who received
glyceryl trinitrate through the PVC set had the same clinical response as
patients who had received the drug through the polyethylene set and there
were no differences between the initial dosage adjustment time, the number of
dosage adjustments and indications for them, the duration of the infusion, the
time of tubing changes and total dose of glyceryl trinitrate.

d) Isosorbide Dinitrate
This long-acting nitrate has been used orally, sublingually and intravenously
in the treatment and prevention of angina and in congestive heart failure
therapy. As stated in its current United Kingdom data sheet, isosorbide
dinitrate is incompatible with PVC containers, but may be used with glass and
polyethylene (ASSOCIATION OF BRITISH PHARMACEUTICAL INDUSTRY 1994).
COS SUM and ROBERTS (1981) investigated the availability of this drug from
intravenous delivery systems. Appreciable loss of the drug was found follow-
ing storage in PVC infusion bags and/or in burettes (cellulose propionate) of
administration sets. This degree of loss was also related to the flow rate of the
infusion; lower rates of recovery of the nitrate were found at slower flow rates.
No loss of isosorbide dinitrate was found during the storage of the drug in glass
for 200h.
SAUTOU et al. (1994) evaluated the compatibility of isosorbide dinitrate
and heparin with polypropylene, polyethylene and PVC under simulated clini-
cal conditions. Results indicated that both drugs were compatible with
polypropylene syringes; preparing the syringes 8h in advance did not alter the
stability of the drugs, provided that the nitrate was not refrigerated. The
concentration of isosorbide dinitrate infused through PVC/polyethylene tub-
ing (PVC on the outside and polyethylene on the inside) was unaffected,
irrespective of whether it was administered alone or together with heparin. As
was expected, both drugs were sorbed by the PVc. The nitrate was rapidly
fixed to the PVC and then released again; this release occurred earlier when
heparin was co-administered. It was concluded by the authors that adsorption
and release of the drugs appeared to be very random.

2. Leaching
The term leaching is applied to the migration of a substance from packaging
material into a medicine. It is influenced in rate and extent by the solvent
system used in the preparation, and pH and temperature conditions during
processing and storage (COOPER 1974). Examples are the leaching of alkali
from soda-glass bottles and the leaching of zinc salts, used as activators, from
rubber closures. Barium ions leached from borosilicate glass may react with
sulphates of drugs such as kanamycin, atropine or magnesium to form barium
In Vitro Drug Interactions 263

sulphate crystals; the presence of sulphurous acid salts as antioxidants in

solutions may result in the formation of crystals (LUND 1994).
Leaching of plastics components has received a great deal of attention,
especially in relation to parenteral preparations. MOORHATCH and CHIOU
(1974) examined the leaching of chemicals from PVC bags containing various
solutions. Distilled water, normal saline, 5% dextrose solution, 5% alcohol
solution and aqueous buffers produced insignificant leaching from PVC bags.
Solutions of the surfactants Tween 20 and Tween 80 produced significant
leaching of the plasticiser diethylhexyl phthalate (DEHP) from PVC bags.
Leaching was found to increase with the surfactant concentration and storage
time. VENKATARAMANAN et al. (1986) studied the leaching of DEHP from
flexible PVC containers into intravenous cyclosporin A solutions. As found in
the MOORHATCH and CHIOU study (1974), the amount of DEHP leached into
solutions was dependent on storage time. By 48h, 33mg DEHP had leached
into the solution. The leaching effect was attributed to the presence of a non-
ionic surfactant, polyoxyethylated castor oil, the injection vehicle for
cyclosporin A. PEARSON and TRISSEL (1993) found that DEHP was leached
from PVC containers by polysorbate 80, polyoxyethylated castor oil,
cyclosporin, miconazole, teniposide, chlordiazepoxide hydrochloride,
etoposide and the vehicles used in paclitaxel and taxotere formulations. The
importance of the paclitaxel vehicle as a causative agent of leaching was
further confirmed by a later study (TRISSEL et al. 1994). Paclitaxel, used in
metastatic ovarian cancer, is formulated in 49% dehydrated alcohol and
polyoxyethylated castor oil, the latter being known to leach the plasticiser
(DEPH) from PVC infusion containers and intravenous administration sets
(PEARSON and TRISSEL 1993). It is recommended that paclitaxel is not to be
used with polyvinyl chloride equipment; a short PVC inlet or outlet on a filter
may be acceptable (BRITISH NATIONAL FORMULARY 1995). Glass, polypropylene
and polyolefin containers are recommended for use with this formulation. This
recommendation was largely supported by TRISSEL'S study, in which the com-
patibility of the paclitaxel injection vehicle (drug was not used), corresponding
to paclitaxel injection 0.3 and 1.2mg/ml, was evaluated with a variety of
intravenous administration and extension sets by measuring the amount of
DEHP leached from the sets (TRISSEL et al. 1994). All the extension sets were
compatible with the injection vehicle; however, two administration sets were
incompatible with the vehicle simulating the low concentration (0.3 mgl/ml)
injection, and five sets were incompatible with the vehicle simulating the high
concentration (1.2mg/ml) injection. In addition, leaching of DEHP occurred,
despite some of the incompatible administration sets being labelled as not
containing PVc. Careful selection of administration and extension sets was
suggested as one way of overcoming potential leaching when using paclitaxel
injection.
In earlier work, ALLWOOD (1986) demonstrated that the Intralipid compo-
nent of a TPN mixture could also extract DEHP from PVC administration
sets. In this study, the sets were primed according to the manufacturer's
recommendations and contained the fat emulsion alone (10% or 20%), a TPN
264 J.e. McELNAY and C.M. HUGHES

regimen containing fat emulsion (20%) or a TPN regimen without added

Intralipid; storage was at 4°_6°C or at ambient temperature for 24h. Extrac-
tion depended on the storage temperature; refrigeration (4°-6°C) reduced the
leaching of DEHP. Slightly more DEHP was extracted into the 10% com-
pared to 20% Intralipid, although the difference was not significant. The rate
of extraction of the plasticiser into a TPN regimen containing the fat emulsion
was substantially lower than into undiluted Intralipid; quantities released into
the TPN regimen without Intralipid were negligible. It was suggested that
patients receiving Intralipid infused through a standard PVC-containing ad-
ministration set would receive a small dose of DEHP. Although it was consid-
ered that this would not be a problem in most patients, pregnant mothers and
neonates would be at greater risk and, thus, it is now generally accepted that
non-DEHP-containing sets (made from ethyl vinyl acetate) should be used to
administer Intralipid-containing TPN fluids.

3. Permeation
Permeation is another physicochemical mechanism of drug binding associated
with some forms of pharmaceutical packaging. It is a combination of absorp-
tion followed by migration to the outer surface, from which the drug is
released by evaporation. Losses may be substantial and they continue
throughout the period of administration since saturation of the plastic is never
achieved (ALLWOOD 1994). Permeation may also occur due to passage of
gases, vapours and liquids through packaging; water vapour permeates
through silicone rubber, and to varying degrees through plastics, particularly
PVC and polystyrene (LUND 1994).
In a study conducted by KOWALUK et al. (1981) a number of hypnotic
agents were investigated in relation to incompatibility with PVC infusion bags.
Chlormethiazole and diazepam were found to be substantially lost after 24h of
storage (33% and 20%, respectively). The chlormethiazole was shown to
penetrate through the plastic as evidenced by a distinctive odour of the drug in
the immediate vicinity of the bag. COS SUM and ROBERTS (1981) also reported
a loss in drug concentration when chlormethiazole was stored and infused
through PVC bags and infusion sets; the average loss from the infusion fluid
approximated to 15%-30%. Loss was negligible when infused from glass
syringes through polyethylene tubing, thereby presenting an alternative to
PVC materials. As mentioned above the loss of glyceryl trinitrate from intra-
venous fluid containers involves a degree of permeation through plastic mate-
rials followed by loss by evaporation (particularly in the case of PVC). This
phenomenon is not limited to intravenous preparations of the drug, i.e., sublin-
gual glyceryl trinitrate should be stored in glass bottles.

4. Polymer Modification
In the work of KOWALUK et al. (1981) described in the previous section, it was
also reported that the PVC infusion bags containing chlormethiazole became
In Vitro Drug Interactions 265

more pliable and softer, particularly at higher concentrations of the drug. It

would seem that chlormethiazole may be responsible for this effect, rather
than a surfactant or cosolvent as it is formulated in an aqueous vehicle. It was
suggested that at high concentrations chlormethiazole caused greater PVC
polymer-chain flexibility through plasticisation. In this way, it enhanced its
own sorption and permeation through the plastic. Such effects were partly
attributed to its high lipid solubility.

IV. Drug Interactions with Contact Lenses

Although this phenomenon is somewhat different to the interactions already
discussed, it is still appropriate to discuss drug interactions with contact lenses
in this section. A number of drugs are known to cause problems for those who
wear contact lenses. Reports have suggested that some drugs can enter into
surface interactions with the lens: rifampicin is associated with a permanent
and damaging discoloration of the lens plastic (LYONS 1979). Adenochrome
pigmentation associated with the use of adrenaline eye drops has been re-
ported, involving the hydrophilic lenses of three elderly patients; in each case
the lens had dense brown coloration (SUGAR 1974). Sulphasalazine used in the
treatment of colitis has been implicated in the staining of a soft contact lens
(RILEY et al. 1986). The patient wore the soft contact lens in one eye and a gas-
permeable lens in the other. During the first 48 h of sulphasalazine treatment,
the lenses were not worn; after this time, the lenses were refitted and yellow
staining of the soft lens was noted. It is now commonplace for those patients
undergoing topical treatment with eye drops and who also wear lenses to be
advised to remove their lenses for the duration of treatment, unless eye drops
are specifically indicated as safe to use with contact lenses. It is more difficult,
however, to predict possible interactions in those patients undergoing systemic
therapy, and it is only as a result of isolated reports in the literature that
clinicians are in a position to advise patients accordingly.

D. In Vitro Drug Interactions

in Therapeutic Drug Monitoring and Drug Analysis
The examples discussed above represent interactions which may occur in the
clinical setting. However, there are also examples of in vitro drug interactions
occurring in the laboratory setting which may have implications for thera-
peutic drug monitoring and drug assays, and in turn may have clinical
consequences.

I. Cyclosporin
This cyclic peptide is widely used in organ and tissue transplantation and is
associated with a number of side effects, most notably nephrotoxicity (CALNE
266 J.e. McELNAY and e.M. HUGHES

et al. 1978). This has led to the development of a number of therapeutic

monitoring techniques, e.g., radioimmunoassay (RIA) and high-performance
liquid chromatography (HPLC). However, a number of these techniques gave
erroneous cyclosporin levels due to analytical problems (SHAW et al. 1987).
The use of an inappropriate pipette for manipulating cyclosporin standards
has led to low measured concentrations of the drug. This was due to the
pipette having a plastic tip (rather than a glass tip) to which cyclosporin
adsorbed. Sorption was attributed to the lipophilic nature of cyclosporin
(McMILLAN 1989) and it is now recommended that serum level monitoring
techniques should use glass where possible.

II. Chloroquine
GEARY et al. (1983) showed that in the laboratory setting preparations of
chloroquine in various solutions showed decreases in concentration of up to
40% when stored in glass containers and that the drug was extensively bound
to cellulose acetate filters. YAHYA et al. (1985) examined the binding of
chloroquine to different grades of laboratory glassware at different concentra-
tions and different pH conditions. Storage in soda glass (test tubes and glass
wool) showed a decrease in original drug concentration of up to 60% and 97%,
respectively. The highest degree of binding was recorded at physiological
pH (7.4) and at low concentrations. Borosilicate glass did not exhibit any
chloroquine binding. A second study investigated the binding of this anti-
malarial agent to a range of plastic materials under conditions of varying pH
and concentration (YAHYA et al. 1986). Chloroquine was sorbed by cellulose
propionate, ethylvinyl acetate, methacrylate butadiene styrene, polyethylene,
polypropylene and PVC, but not by high-density polystyrene. Passing
chloroquine solutions through cellulose acetate filters once resulted in a high
loss of drug. The unionised form of the drug was preferentially sorbed. These
studies indicate that care must be taken when quantifying chloroquine in the
laboratory since the drug will come into contact with a range of materials. This
is particularly the case in kinetic studies and malarial sensitivity testing, when
chloroquine may be bound to plastic syringes or pipettes during sampling;
binding to membrane filters used to clarify and/or to sterilise chloroquine
solutions prior to use or analysis could also give rise to erroneous results
(YAHYA et al. 1986). Although this problem has been highlighted for
chloroquine, it is well known that care must be taken in the manipulation of
other drugs in the laboratory setting, e.g., some benzodiazepines are highly
bound to glassware and this necessitates the silylation of glassware prior to
use.

E. In Vitro Drug Interactions

as a Result of Formulation Changes
The final category of in vitro drug interactions which will be discussed in this
chapter are those which occur as a result of changes in a product formulation;
In Vitro Drug Interactions 267

this may result in changes in bioavailability of the active drug constituent and
there have been a number of reported cases where this has led to adverse
clinical consequences for patients (McELNAY and D' ARCY 1980). Other re-
ports are from experimental studies which stress the importance of formula-
tion in relation to bioavailability.
The two classical examples of this type of in vitro effect involve the drugs
phenytoin and digoxin. Anticonvulsant intoxication resulted from subtle
changes to the formulation of phenytoin capsules (TYRER et al. 1970); lactose
replaced calcium sulphate as an excipient, resulting in increased dissolution
and higher serum levels of phenytoin as illustrated in Fig. 5. Particle size
changes to digoxin also led to marked changes in digoxin levels (JOHNSON et al.
1973). Tablets manufactured after May 1972 contained a smaller particle size
digoxin, which resulted in a higher degree of absorption of the drug and
overdigitilisation of patients. This led to much stricter statutory controls in
relation to formulation in developed countries. However, a later study by
FLETCHER and SUMMERS (1980) revealed that these controls did not extend to
all countries. Five brands of digoxin, which were available in South Africa at
the time, were compared in relation to physical and dissolution characteristics
and bioavailability. Only one brand satisfied all requirements, based on official
specifications. One brand did not satisfy diameter specifications, was soft and
crumbled easily, and showed only 34.82% dissolution within the period exam-
ined. A second brand was viewed as a very low grade product which neither
dispersed readily nor complied with the specifications for digoxin content. It is
this type of report which has given rise to concerns over the quality of some
generic products available in developing countries. For most products, this is

Phen'1toin Phenytoin Phenytoin

(lactose) ((0504 ) (lactose)

~ ~
30 "T1

-:25
.....
I
--
E -0
i"
[ 20
c:
~:i"
o.§ 15 30 CD
>-
c: ~
II iii
20 5-
...
~IO
::J
8 5 10
'3
<= «&
0 0
!
8 16 2'1 32 o 8
Doy
Fig. 5. Blood phenytoin concentrations in a patient taking phenytoin (400mglday),
with excipients respectively as shown (lactose, calcium sulphate, lactose). Vertical
columns represent daily faecal excretion of phenytoin when measured. (After TYRER
et al. 1970)
268 J.e. McELNAY and C.M. HUGHES

not a problem and slight changes in bioavailability are of no clinical impor-

tance; however, many authorities recommend that products such as phenytoin
and lithium, which have a narrow therapeutic index, should be prescribed by
proprietary (brand) name to ensure that the patient receives the same formu-
lation each time.

I. Problems with Sustained-Release Formulations

One other drug for which prescription by brand name is strongly recom-
mended and whose formulations and kinetics has been widely investigated is
theophylline. WEINBERGER et al. (1978), in a study of the relation of product
formulation to absorption, found that the absorption of theophylline from
both a solution and uncoated tablets was rapid and complete. Three of six
sustained-release formulations were completely and consistently absorbed,
although the absorption was slower, as would be expected, than with the
solution and uncoated tablets. Absorption of theophylline from the other
three sustained-release products appeared to be more erratic and less com-
plete as shown in Fig. 6. Further work by this group examined the absorption
characteristics of a once-daily dosage formulation, Theo-24, compared to a
theophylline reference solution, in healthy volunteers (HENDELES et al. 1985).
Absorption of Theo-24 after an overnight fast was very slow, with only 71 % ±
6% (±SE) of the dose ultimately absorbed. In contrast, food caused precipi-
tous "dose-dumping", resulting in peak levels of drug in the serum that aver-
aged 2.3 times higher than after a fasting dose. About half of the dose was
absorbed in a 4-h period, generally beginning 6-8h after postprandial admin-
istration, and complete absorption was then attained within 24h. Theophylline
toxicity was evident in four subjects when they took the dose postprandially
whereas no toxic effects occurred during the fasting regimen. Although this
example is not strictly an in vitro drug interaction as previously defined, it does
illustrate the influence of formulation on clinical outcome. The pattern of
absorption following breakfast was consistent with a greatly increased dissolu-
tion of the formulation (encapsulated coated beads which are designed to
dissolve slowly) following exposure to the alkaline bile salts or pancreatic
enzymes (or both) that are secreted following post-prandial gastric emptying.
Theophylline is available worldwide as a number of modified-release for-
mulations, e.g., Nuelin, Slo-Phyllin, Theo-Dur and Uniphyllin. Despite a wide
availability of branded products, work is still continuing on the development
of other modified-release theophylline preparations. Modification of the for-
mulation components gives rise to changes in the dissolution, binding to and
release from different formulations. Numerous publications in the pharmaceu-
tical literature explore the influence of formulation on drug release and conse-
quent bioavailability. Illustrative of this work is a recent study by DILucCIO et
al. (1994), who examined crystalline polyvinyl alcohol (PVA) polymer and
low-crystallinity polyvinyl alcohol-methyl acrylate copolymer (PVA-MA) as
sustained-release excipients with theophylline. By blending of different pro-
In Vitro Drug Interactions 269

1.0
.9
.8

.7
.6

.5
Theophylline Solullon

.4 SUSTAINED RELEASE BEAD

FILLED CAPSULES
.3 6----4 Siophyilin Gyrocapo N~12

_-_TheophyISR N~14
.2 cr - ~ Aerolate N ~4
.---.. Theobld N~6
.1
O~-L~~-L--~~~~~~~~~~~--
o 4 6 8 10 12 14 16 18 20
a Time (hours)

::~ :r-~==I
I
1.0 ................... T .. __ ...
,-.---- ----. 1
.9 ".1_+-.1':::-1-'-1
.8
1''' ,,1--1
I .

..
1)
III
.Q .7 /Y'{
0
~),/
ll·
CI)
.Q .6
0:(

'-e-
c: .5
0
( ,)
.4
t;f. Theophylline Solullon

r~
4-
.3 SUSTAINED RELEASE TABLETS
. - ... Theodur-l00mg/ N~5
.2 t{ 0 - - ~ Theodur· 300mg I
..---. Aminodur N::5

.1

0
2 4 6 8 10 12 14 16 18 20 22 24 26 28
b Time (hours)
Fig. 6. a Absorption of sustained-release bead-filled capsules of theophylline. b Ab-
sorption of sustained-release tablets of theophylline. Aminodur, Aerolate and Theobid
are seen to have suboptimal absorption. (After WEINBERGER et al. 1978)
270 J.e. McELNAY and e.M. HUGHES

portions of PVA and PVA-MA it was possible to affect the release character-
istics of the drug. Tablets made with crystalline PVA provided instant release
of theophylline in vitro, whereas those made with a larger proportion of PVA-
MA relative to PVA demonstrated a very prolonged release profile. It was
concluded that the crystallinity of the polymer played a major role in the
release of the drug, as high-crystallinity polymers (PVA) are less permeable to
diffusion than low-crystallinity polymers (PVA-MA); the latter are known to
dissolve more slowly in water and hence drug release is sustained. The
bioavailability (in dogs) of a formulation containing PVA:PVA-
MA: theophylline in the ratio of 1: 9: 10 was equivalent to that obtained after
administration of a commercially available product Theo-Dur.
As previously stated, theophylline is a drug which is normally prescribed
by brand name to ensure that the patient receives the same release/absorption
patterns of the drug each time it is administered. A recent study by HENDELES
et al. (1995) evaluated the relative bioavailability and clinical efficacy of a
generic slow-release theophylline tablet and Theo-Dur in 14 adults with
asthma. The results of the study indicated that the generic slow-release theo-
phylline tablet and Theo-Dur were not bioequivalent. Although they were
absorbed to the same extent, as indicated by similar areas under the serum
concentration-time curves and peak concentrations, the generic product was
absorbed more rapidly. This was demonstrated by the shorter time to peak
concentration, larger fluctuations between peak and trough concentrations
and slightly lower mean trough concentration. Therapeutic equivalence was
also assessed using attenuation of the response to exercise as a surrogate for
clinical effect (exercise challenge was 17 h after last dose of theophylline).
Although neither product effectively attenuated exercise-induced
bronchospasm, the authors considered that despite the differences in absorp-
tion rates both formulations were therapeutically equivalent, but not
bioequivalent.
As a result of stricter statutory controls with regard to changes in product
formulation, there are now relatively few clinical reports of drug interactions
arising from formulation modification.

II. Drug Excipient Interactions

The BRITISH PHARMACEUTICAL CODEX (LUND 1994) documents a number of
drug-excipient interactions. Ionic interactions between cationic and anionic
entities may be particularly problematic. Carmellose sodium, an anionic poly-
mer, reacts with large cations such as neomycin and chlorpromazine in solu-
tion; formation of a precipitate depends upon the conditions and also the
concentrations of each substance. Benzalkonium chloride and other cationic
antimicrobial preservatives are inactivated to varying degrees in the presence
of carbomer and other anionic polymers. Complex formation between drug
and excipient may also be common. Phenolic preservatives may be partly
inactivated in complexes formed with macrogol derivatives. The non-ionic
In Vitro Drug Interactions 271

excipient povidone may complex with anionic or cationic dyes, and drugs such
as chlorpromazine and chloramphenicol. Gels made with povidone may in-
crease in viscosity as a result of complexation with 8-hydroxyquinolone
sulphate or thiomersal sodium. Excipients (or another drug) may also affect
the stability of a drug, leading to reduced bioavailability and changed thera-
peutic efficacy as reported by the BRITISH PHARMACEUTICAL CODEX (LUND
1994): sodium metabisulphite rapidly inactivates cisplatin and propylene gly-
col and macro go Is catalyse the degradation of benzoyl peroxide to benzoic
acid and carbon dioxide.
An early study by MCGINITY and LACH (1976), using dissolution and
dialysis techniques, examined the adsorption of various drugs to montmorillo-
nite clay; this colloidal magnesium aluminium silicate clay has been used as a
disintegrant, binder and lubricant. Chlorpheneramine maleate, amphetamine
sulphate and propoxyphene hydrochloride were all found to be strongly
bound to montmorillonite due to their cationic nature; amphoteric compounds
such as theophylline and caffeine were found to bind moderately to the clay,
whereas sodium salicylate and sulphanilamide (anionic in nature) were re-
leased rapidly from montmorillonite, i.e., binding was minimal.
Although these latter binding interactions are largely regarded as detri-
mental in that drug absorption will be reduced, the binding can be used to
advantage in the formulation of sustained-release products. The interaction of
piroxicam, for example, with various synthetic polymers (polyethylene glycol
400,600,1000,4000,6000 and 20000, polyvinylpyrrolidone 30000, 40000 and
64000, and dextran 40000 and 75000) has been determined by ELSHAITAWY et
al. (1994). The binding of piroxicam in different concentrations with the differ-
ent polymers showed that increasing drug concentration was accompanied by
an increase in the amount of bound drug to the polymer, until near saturation.
The binding capacity of piroxicam was found to be greatest for the poly-
vinylpyrrolidones and least for dextran.
CHOWHAN and CHI (1986a) compared the two lubricants magnesium stear-
ate and sodium stearyl fumarate under identical conditions, in order to study
their roles in drug-excipient interactions; the drug in question was ketorolac.
After prolonged mixing, sodium stearyl fumarate did not affect the drug of
excipient (crospovidone) and, as a result, the disintegration time and drug
dissolution from hand-filled capsules were not adversely affected. In contrast,
magnesium stearate did demonstrate a drug-crospovidone interaction which
resulted in an increased disintegration time and dissolution time. This resulted
from particle-particle interaction whereby lamination and flaking of magne-
sium stearate occurred. These flakes adhered to drug particles and caused a
significant reduction in the dissolution rate. In a further study, using
prednisone, the effect of powder mixing on drug-excipient interactions and
their effect on in vitro dissolution from capsules was studied (CHOWHAN and
CHI 1986b). Two powder formulations contained dibasic calcium phosphate
dihydrate as a filler and potato starch or sodium starch glycolate as a
disintegrant were studied. The third powder formulation contained
272 J.e. McELNAY and e.M. HUGHES

pregelatinised starch as a disintegrant/filler. The lubricant in all formulations

was magnesium stearate. When prednisone, calcium filler and disintegrant
were mixed thoroughly and hand-filled into capsules, only 70% of the drug
dissolved in 30min. The decrease in the rate of drug dissolution resulted from
formation of agglomerates (calcium filler and disintegrant) and the inclusion
of drug particles in these agglomerates. When a mixture of prednisone and
pregelatinised starch was used, complete dissolution of the drug was achieved
after 30min; formation of agglomerates did not occur with this formulation,
and complete dissolution resulted.
The compatibility of pyridoxine hydrochloride (vitamin B6) with various
tableting excipients has been investigated using a variety of techniques, e.g.,
isothermal stress testing and differential scanning calorimetry (DURIG and
F ASSIHI 1993). The experiments revealed chemical incompatibilities between
the drug and mannitol, lactose and corn starch. Various cellulose derivatives
and colloidal silicone dioxide were identified as strong stabilising factors in
formulations. It appeared that these excipients acted as moisture scavengers,
thereby decreasing the amount of free water available for interaction with
pyridoxal hydrochloride.
In a more recent study, cellulose derivatives were found to have an effect
on the stability of diethylpropion both in terms of stability and bioavailability
(GOMEZ et al. 1993). In vitro results showed that sodium carboxymethylcellu-
lose and methylcellulose caused a marked degradation of diethylpropion.
Dissolution assays revealed that the availability of the drug was also vastly
reduced by these cellulose derivatives.

III. Formulatiou Effects on Rectal Bioavailability

Bioavailability/formulation interactions are not restricted to those prepara-
tions which are administered orally. This was illustrated by McELNAY and
NICHOL (1984), who compared the performance of a novel flow-through bead-
bed dissolution apparatus for suppositories with that of the more traditional
rotating basket technique. Similar dissolution profiles were obtained for fatty-
based benzocaine suppositories using the two methods at 37°C. A second set
of experiments measured the dissolution profiles, using these two techniques,
of two commercially available indomethacin suppository formulations
(Indocid and Imbrilon). Although both products contained the same dose of
indomethacin in a polyethylene glycol base, clear differences were noted
between the dissolution profiles of the two products; a markedly decreased
rate of dissolution was noted for Indocid suppositories and this difference was
particularly marked when using the flow-through bead-bed apparatus (Fig. 7).
A later study examined the serum indomethacin profiles after rectal adminis-
tration of these two brands of indomethacin in healthy volunteers (McELNAY
et al. 1986). Statistical analysis indicated that there were no significant differ-
ences between the two formulations in relation to any of the pharmacokinetic
parameters which were compared, although the time taken to attain maximal
In Vitro Drug Interactions 273

100

90

80

"0 70
OJ
>
~ 60
.!!l
"0
Ol 50

"
:J

'0 40
OJ
g> 30
C
OJ
(.)
Q; 20
11.

10

30 60 90 120 150 180

a Dissolution time (min)

100

90

80

"0 70
OJ
>
(560
l:l
:0
Ol
50
2
~o 40
OJ
g> 30
C
OJ
(.)

30 60 90 120 150 180

b Dissolution time (min)

Fig. 7. a Dissolution profiles of indomethacin suppositories obtained using the flow-

through bead-bed apparatus: ., Indocid; 0, Imbrilon. Mean ± SE data presented. b
Dissolution profiles of indomethacin suppositories obtained using the rotating basket
apparatus: ., Indocid; 0, Imbrilon. Mean ± SE data presented. (After McELNAY and
NICOL 1984)
274 J.e. McELNAY and e.M. HUGHES

plasma concentration was somewhat greater with the Indocid suppositiories

(1.75 ± O.16h) than with the Imbrilon formulation (1.42 ± O.17h).
MORGAN et al. (1992) investigated the release of morphine (15mg) from a
number of suppository bases using the United States Pharmacopeia rotating
basket dissolution apparatus. Morphine hydrochloride in polyethylene glycol
(a hydrophilic suppository base), morphine alkaloid in polyethylene glycol,
and morphine hydrochloride in Novata BBC (a lipophilic base) completely
released the drug within 25 min, whereas morphine alkaloid in Novata BBC
released the drug over 10 h. This prolonged release was also seen in vivo when
all formulations were used in a clinical trial; morphine alkaloid in Novata BBC
was viewed as a useful long-acting preparation of the opioid and warranted
further investigation. Clearly, the choice of suppository base was critical in
relation to bioavailability of the drug.

F. Conclusions
Interpretation of in vitro drug interactions involves a number of basic prin-
ciples (ALLwooD 1994):
1. An understanding of the chemistry of drugs in the salt and/or free form and
the effects of pH
2. A knowledge of the basis for physicochemical interactions between solutes
and solids
3. Implementing practical solutions to overcome an in vitro drug interaction
4. Evaluation of the clinical effect of the in vitro drug interactions
Many of the interactions which have been discussed in this chapter may be
explained in chemical and physical terms. Once the interaction has been
elucidated a number of approaches can be used to overcome or avoid un-
wanted sequelae. Ways of overcoming the interactions include:
1. Avoidance of direct contact of interacting agents, e.g., via use of multiuse
cannulae for intravenous administration
2. Adding increased doses of drugs to infusion fluids when the drug is bound
to the infusion apparatus
3. Ensuring that oral formulations are adequately tested for bioavailability
prior to clinical use
4. Use of glass or suitable plastics which do not interact with drugs in question,
e.g. use of polypropylene-lined infusion bags and administration sets for
drugs which are highly bound to PVC

References
Airaudo CB, Gaytesorbier A, Bianchi C, Verdier M (1993) Interactions between six
psychotherapeutic drugs and containers: influence of plastic material and infusion
solutions. Int J Clin Pharmacol Ther 31:261-266
In Vitro Drug Interactions 275

Allwood MC (1986) The release of phthalate ester plasticizer from intravenous admin-
istrations sets into fat emulsion. Int J Pharm 29:233-236
Allwood MC (1994) Drug stability and intravenous administration, 2nd edn. Clinical
pharmacy practice guide. UKCPA, University of Leeds Media Service, Leeds
Altavela JL, Haas CE, Nowak DR, Powers J, Gacioch GM (1994) Clinical response to
intravenous nitroglycerin infused through polyethylene or polyvinyl chloride tub-
ing. Am J Hosp Pharm 51:490-494
Association of British Pharmaceutical Industry (ABPI) (1994) Data sheet compen-
dium 1994/95. Datapharm, London
Autain J (1963) Plastics in pharmaceutical practice and related fields, part I. J Pharm
Sci 52:1-23
Baaske DM, Amann AH, Wagenknecht DM, Mooers, M, Carter JE, Hoyt HJ, Stoll
RG (1980) Nitroglycerin compatibility with intravenous fluid filters, containers,
and administration sets. Am J Hosp Pharm 37:201-205
Benvenuto JA, Anderson RW, Kerkof K, Smith RG, Loo TL (1981) Stability and
compatibility of antitumor agents in glass and plastic containers. Am J Hosp
Pharm 38:1914-1918
Birely A W, Scott MJ (1982) Plastic materials, properties and applications. Blackie,
Glasgow, pp 1-18
Bonhomme L, Benhamou D, Beugre T (1992) Slow-release effect of pH-adjusted
bupivacaine: in vitro demonstration. Int J Pharm 84:33-37
British National Formulary (1994) No 28. British Medical Association and Pharmaceu-
tical Press, London
British National Formulary (1995) No 29. British Medical Association and Pharmaceu-
tical Press, London
Brydson JA (1982) Plastics materials. Butterworth, London, pp 18-143
CaIne RY, White DRG, Thiru S (1978) Cyclosporin A in patients receiving renal
allografts from cadaver donors. Lancet 11:515-524
Carter SJ (1975) Cooper and Gunn's dispensing for pharmaceutical students, 12th edn.
Pitman, London, pp 253-264
Chen ML, Chiou WL (1982) Adsorption of methotrexate onto glassware and syringes.
J Pharm Sci 71:129-131
Chiou WL, Moorhatch P (1973) Interaction between vitamin A and plastic intravenous
bags. JAMA 223:328
Chowhan ZT, Chi LH (1986a) Drug-excipient interactions resulting from powder
mixing. IV. Role of lubricants and their effect on in vitro dissolution. J Pharm Sci
75:542-545
Chowhan ZT, Chi LH (1986b) Drug-excipient interactions resulting from powder
mixing. III. Solid state properties and their effect on drug dissolution. J Pharm Sci
75:534-541
Collins JL, Lutz RJ (1991) In vitro study of simultaneous infusion of incompatible
drugs in multilumen catheters. Heart Lung 20:271-277
Cooper J (1974) Plastic containers for pharmaceuticals testing and control. WHO,
Geneva, pp 1-13
Cossum PA, Roberts MS (1981) Availability of isosorbide dinitrate, diazepam and
chlormethiazole from i.v. delivery systems. Eur J Clin PharmacoI19:181-185
DiLuccio RC, Hussain MA, Coffin-Beach D, Torosian G, Shefter E, Hurwitz AR
(1994) Sustained-release oral delivery of theophylline by use of polyvinyl alcohol
and polyvinyl alcohol-methyl acrylate polymers. J Pharm Sci 83:104-106
Drott P, Meurling S, Meurling L (1991) Clinical adsorption and photodegradation of
the fat-soluble vitamin A and vitamin E. Clin Nutr 10:348-351
Durig T, Fassihi AR (1993) Identification of stabilizing and destabilizing effects of
excipient-drug interactions in solid dosage form design. Int J Pharm 97:161-170
Elshattawy HH, Ahmed EI, Bayomi MA, Abdelfatah 1M (1994) Piroxicam synthetic
polymers interactions. Pharm Ind 56:396--399
Ferrebee JW, Johnson BB, Mithoefer JC, Gardella JW (1951) Insulin and adrenocor-
ticotropin labelled with radio-iodine. Endocrinology 48:277-283
276 J.e. McELNAY and e.M. HUGHES

Fletcher S, Summers RS (1980) Physical and dissolution characteristics and

bioavailability of five brands of digoxin tablets. S Afr Med J 57:530-533
Fournier C, Hecquet B, Bastian G, Khayat D (1992) Modification of the physicochemi-
cal and pharmacological properties of anticancer platinum compounds by com-
mercial 5-fluorouracil formulations - a comparative study using cisplatin and
carboplatin. Cancer Chemother. Pharmacol 29:461-466
Geary TG, Akood MA, Jensen JB (1983) Characteristics of chloroquine binding to
glass and plastic. Am J Trop Med Hyg 32:19-23
Griffin JP, D' Arcy PF (1988) Drug interactions in vitro. In: A manual of adverse drug
interactions, 4th edit. Wright, Bristol, pp 6-16
Gomez M, Alvarez A, Cid E (1993) Influence of cellulosic derivatives on in vitro
stability and availability of diethylpropion in prescription orders. Rev Med Chile
121:1013-1016
Hansen HC, Spillum A (1991) Loss of nitroglycerin during passage through two differ-
ent infusion sets. Acta Pharm Nord 3:131-136
Hendeles L, Weinberger M, Milavetz G, Hill M, Vaughan L (1985) Food-induced
"dose-dumping" from a once-a-day theophylline product as a cause of theophyl-
line toxicity. Chest 87:758-765
Hendeles L, Breton AL, Beaty R, Harman E (1995) Therapeutic equivalence of a
generic slow-release theophylline tablet. Pharmacotherapy 15:26-35
Hill JB (1959) Adsorption of insulin to glass. Proc Soc Exp BioI Med 102:75-77
Holmes SE, Aldous S (1991) Stability of miconazole in peritoneal dialysis fluid. Am J
Hosp Pharm 48:286-290
Jenke DR (1993) Modelling of solute sorption by polyvinyl chloride plastic infusion
bags. J Pharm Sci 82:1134--1139
Johnson BF, Fowle ASE, Lader S, Fox J, Munro-Faure AD (1973) Biological activity
of digoxin from Lanoxin produced in the United Kingdom. Br Med J 4:323-326
Kowaluk EA, Roberts MS, Blackburn HD, Polack AE (1981) Interactions between
drugs and polyvinyl chloride infusion bags. Am J Hosp Pharm 38:1308-1314
Lee MG (1986) Reduced absorption of drugs onto polybutadiene administration sets.
Am J Hosp Pharm 43:1945-1950
Li LC, Zhang H, Pecosky DA (1993) The effect of surfactant on drug-plastic sorption
phenomenon in solution. J Pharm PharmacoI45:748-749
Loucas SP, Maager P, Mehl B, Loucas ER (1990) Effect of vehicle ionic strength on
sorption of nitroglycerin to a polyvinyl chloride administration set. Am J Hosp
Pharm 47:1559-1562
Lund W (ed) (1994) British pharmaceutical codex. Pharmaceutical Press, London, pp
311-322
Lyons RW (1979) Orange contact lenses from rifampicin. N Engl J Med 300:372-373
Malick A W, Amann AH, Baaske DM, Stoll RG (1981) Loss of nitroglycerin from
solutions to intravenous plastic containers: a theoretical treatment. J Pharm Sci
70:798-800
McElnay JC, D'Arcy PF (1980) Sites and mechanisms of drug interactions. I. In vitro,
intestinal and metabolic reactions. Int J Pharm 5:167-185
McElnay JC, Nichol AC (1984) The comparison of a novel continuous flow dissolution
apparatus for suppositories with the rotating basket technique. Int J Pharm 19:89-
96
McElnay JC, Taggart AJ, Kerr B, Passmore P (1986) Comparison of serum
indomethacin profiles after rectal administration of two brands of indomethacin
suppositories in healthy volunteer subjects. Int J Pharm 33:195-199
McElnay JC, Elliott DS, D'Arcy PF (1987) Binding of human insulin to burette
administration sets. Int J Pharm 36:199-203
McElnay JC, Elliott DS, Cartwright-Shamoon J, D'Arcy PF (1988) Stability of
methotrexate and vinblastine in burette administration sets. Int J Pharm 47:239-
247
McGinity JW, Lach JL (1976) In vitro adsorption of various pharmaceuticals to mont-
morillonite. J Pharm Sci 65:896-902
In Vitro Drug Interactions 277

McMillan M (1989) Clinical pharmacokinetics of cyclosporin. Pharmacol Ther 42:135-

156
Moorhatch P, Chiou WL (1974) Interactions between drugs and plastic intravenous
fluid bags. II. Leaching of chemicals from bags containing various solvent media.
Am J Hosp Pharm 31:149-152
Morgan DJ, McCormick Y, Cosolo W, Roller L, Zalcberg J (1992) Prolonged release
of morphine alkaloid from a lipophilic suppository base in vitro and in vivo. Int J
Clin Pharmacol Ther 30:576-581
Newerly K, Berson SA (1957) Lack of specificity of insulin 1-131 binding by isolated rat
diaphragm. Proc Soc Exp BioI Med 94:751-755
Parker W A, McCara ME (1980) Compatibility of diazepam with intravenous fluid
containers and administration sets. Am J Hosp Pharm 37:496-500
Pearson SD, Trissel LA (1993) Leaching of diethylhexyl phthalate from polyvinyl
chloride containers by selected drugs and formulation components. Am J Hosp
Pharm 50:1405-1409
Petty C, Cunningham NL (1974) Insulin adsorption by glass infusion bottles,
polyvinyl chloride infusion containers and intravenous tubing. Anesthesiology
40:400-404
Riley SA, Flegg PJ, MandaI BK (1986) Contact lens staining due to sulphasalazine.
Lancet 1:972
Roberts MS, Kowaluk EA, Polack AE (1991) Prediction of solute sorption by polyvi-
nyl chloride plastic infusion bags. J Pharm Sci 80:449-455
Salomies HEM, Heinonen RM, Toppila MAl (1994) Sorptive loss of diazepam, nitro-
glycerin and warfarin sodium to polypropylene-lined infusion bags (Softbags). Int
J Pharm 110:197-201
Sautou V, Chopineau I, Gremeau R, Chevrier R, Bruneaux F (1994) Compatibility
with medical plastics and stability of continuously and simultaneously infused
isosorbide dinitrate and heparin. Int J Pharm 107:111-119
Schildt B, Ahlgren T, Bergham L, Wendt Y (1978) Adsorption of insulin by infusion
materials. Acta Anaesth Scand 22:556-562
Shaw LM, Bowers LD, Demers L, Freeman DJ, Moyer T, Sarghvi A, Seltman H,
Venkataramanan R (1987) Critical issues of cyclosporine monitoring: report of the
Task Force on cyclosporine monitoring. Clin Chern 36:1841-1846
Sugar J (1974) Adenochrome pigmentation of hydrophilic lenses. Arch Ophthalmol
91:11-12
Taormina D, Abdallah HY, Venkataramanan R, Logue L, Burckart GJ, Ptachcinski
RJ, Todo S, Fung 11, Starzl TE (1992) Stability and sorption of FK 506 in 5%
dextrose and 0.9% sodium chloride injection in glass, polyvinyl chloride and
polyolefin containers. Am J Hosp Pharm 49:119-122
Trissel LA, Xu Q, Kwan J, Martinez JF (1994) Compatibility of paclitaxel injection
vehicle with intravenous administration and extension sets. Am J Hosp Pharm
51:2804-2810
Twardowski ZJ, Nolph KD, McGary TJ, Moore HL, Collin P, Ausman RK, Slimack
WS (1983a) Insulin binding to plastic bags: a methodologic study. Am J Hosp
Pharm 40:575-579
Twardowski ZJ, Nolph KD, McGary TJ, Moore HL (1983b) Nature of insulin binding
to plastic bags. Am J Hosp Pharm 40:579-582
Twardowski ZJ, Nolph KD, McGary TJ, Moore HL (1983c) Influence of temperature
and time on insulin adsorption to plastic bags. Am J Hosp Pharm 40:583-586
Tyrer JH, Eadie MJ, Sutherland JM, Hooper WD (1970) Outbreak of anticonvulsant
intoxication in an Australian city. Br Med J 2:271-273
Venkataramanan R, Burckart GJ, Ptachcinski RJ, Blaha R, Logue LW, Bahnson A,
Giam C-S, Brady JE (1986) Leaching of diethylhexyl phthalate from polyvinyl
chloride bags into intravenous cyclosporine solution. Am J Hosp Pharm 43:2800-
2802
Weinberger M, Hendeles L, Bighley L (1978) The relation of product formulation to
absorption of oral theophylline. N Engl J Med 299:852-857
278 J.e. McELNAY and CM. HUGHES: In Vitro Drug Interactions

Wideroe TE, Smeby LC, Berg KL, Jorstad S, Svartas TM (1983) Intraperitoneal (1 251)
insulin absorption during intermittent and continuous peritoneal dialysis. Kidney
Int 23:22-28
Yahya AM, McElnay JC, D'Arcy PF (1985) Binding of chloroquine to glass. Int J
Pharm 25:217-223
Yahya AM, McElnay JC, D'Arcy PF (1986) Investigation of chloroquine binding to
plastic materials. Int J Pharm 34:137-143
Yuen PH, Denman SL, Sokoloski TD, Burkman AM (1979) Loss of nitroglycerin from
aqueous solution into plastic intravenous delivery systems. J Pharm Sci 68:1163-
1166
CHAPTER 10
Age and Genetic Factors in Drug Interactions
J.e. McELNAY and P.F. D'ARCY

A. Introduction
It has long been known that specific patient factors can influence the course of
therapy and the adverse drug reactions and interactions that may follow
(WALLACE and WATANABE 1977; OUSLANDER 1981; BRAVERMAN 1982;
GREENBLATT et al. 1982; SHAW 1982; D'ARCY and McELNAY 1983; ROYAL
COLLEGE of PHYSICIANS 1984; NOLAN and O'MALLEY 1988a,b; DENHAM 1990;
WILLIAMS and LOWENTHAL 1992). Of these factors, age and genetic influences
have been pinpointed as significant contributors to problems with drug
therapy.
This chapter deals with these two influences in turn: Sect. B, "Age", and
Sect. C, "Genetic Factors", and it will highlight some of their more important
influences on drug action. Where possible, it will indicate the mechanisms that
underline these influences.

B. Age
It has long been established that elderly patients use more medicaments than
younger age groups (LANDAHL 1987; WILLIAMS and LOWENTHAL 1992; SLOAN
1992; STEWART and COOPER 1994) and thus have a greater risk of drug-drug
interactions occurring (KELLAWAY and MCCRAE 1973; LAWSON and JICK 1976;
LEVY et al. 1980; WILLIAMSON and CHOPIN 1980; SIMONS et al. 1992; SCHENKER
and BAY 1994; STANTON et al. 1994; STEWART and COOPER 1994) (Fig. 1). It must
be understood, however, that there is virtually no direct evidence in the
literature that age per se can cause drug-drug interactions. Its role is more in
enhancing the effects of drug-drug interactions when they occur. Age thus
tends to exert a quantitative influence on the interaction but does not alter its
qualitative spectrum.
Senescence is frequently evoked to explain the unwanted sequelae to
therapy, undoubtedly rightly - provided senescence is regarded as being ac-
companied, for example, by physiological changes due to age (LAMY 1991), by
small body mass, poor renal function, and impaired function of other organs,
notably the liver. Table 1 shows some of the key changes that occur as a result
of physiological ageing. In elderly patients the reserve capacity of many organs
may be considerably reduced, and because of this erosion there is a narrowing
280 J.e. McELNAY and P.F. D'ARCY

VI 100
U
.!!!
Q;
., 80
.,l!.'
>
"0
co
£: 60
'ji;
VI
C 40
.!!!
iii
.,
0.

Ol
co 20
.,u
C
Q;
[l..

2 3 4 5 6 7 8 9 10

Number of drugs

Fig. 1. Relationship between percentage of patients with ADR and the number of
drugs prescribed. (DENHAM 1990)

of the safety margin between the therapeutic and toxic dose of many drugs. As
a result of this the elderly, as a group, get rather more than their fair share of
drug-induced disease and the complications of drug-drug interactions. This
especially applies to psychogeriatric patients in whom the nature of drugs, the
number used concomitantly, and the long-term use presents peculiar hazards.
Table 2 shows, as an example, the drug-drug interactions identified by
GOSNEY and TALLIS (1984) in their survey of interacting drugs in elderly
patients admitted to hospital. Nine of these combinations were potentially life-
threatening; 51 combinations were likely to lead to potentially serious side
effects, and 27 were significant in as much as they could lead to suboptimal
treatment. Other accounts of common drug-drug interactions are given
throughout this volume and will not be repeated here. It is sufficient to
emphasise that qualitatively the elderly appear to suffer the same adverse
effects as younger age groups but they suffer them at an enhanced level.
Many adverse effects of drugs or drug combinations that may simply be a
nuisance to younger patients are much intensified in the elderly due, for
example, to their decreased resistance to the adverse consequences. Indeed,
the adverse effects of drug treatments may convert the elderly patient from a
functional sentinent human being into a chair-fast incontinent wreck.
In their simplest form, drug-drug interactions cause either an enhanced
pharmacological or toxicological effect of one or more of the combination
drugs, or conversely a reduced therapeutic effect of one or more of the com-
ponents. A number of investigations have shown that age is not an indepen-
dent risk factor (NOLAN and O'MALLEY 1989; CARBONIN et al. 1991; GURWITZ
and AVORN 1991; CHRISCHILLES et al. 1992). However, regardless of whether
the elderly are more sensitive to the adverse effects of medication, they have
more disease, consume more medications and are overrepresented in nearly
every study of adverse drug reactions (SCHNEIDER et al. 1992).
Age and Genetic Factors in Drug Interactions 281

Table 1. Some key changes as a result of physiological aging. (COLLIER et al. 1993)

System Key changes

Cardiovascular Decrease in cardiac output

system Decrease in stroke volume
Loss of elasticity of blood vessels
Decrease in recovery rate to restore basal heart rate
following exercise
Pre-disposition to blood hypercoagulability
Decrease in deposition of fibrin in blood vessel walls
Decrease in peripheral perfusion
Decrease in organ blood flow
Central nervous Changes in receptor and target organ sensitivity
system Decrease in cerebral perfusion
Slowing of cerebral processes
Respiratory system Decrease in the elasticity of lung tissue
Lessening of ciliary activity
Decrease in dead space in lungs
Decrease in forced expiratory volume
Decrease in peak flow
Wasting of muscles in diaphragm
Endocrine system Decrease in glucose tolerance
Decrease in peripheral responsiveness to insulin
Decrease in thyroid functional activity but increased tissue
sensitivity to thyroid hormones
Decrease in normal thyroxine turnover
Reduction of plasma binding of thyroid hormones
Decrease in basal adrenal activity
Reduction of oestrogen level
Decrease in the production of progesterone after the
menopause
Decrease in free androgens in men
Altered hormonal control of bone homeostasis
Compromised immunological systems
Gastrointestinal Decrease in acid secretion
system Decrease in gastric emptying rate
Decrease in gut motility
Decrease in peristaltic activity
Decrease in gastric blood flow
Decrease in absorption of calcium and iron from small
intestine
Hepatic system Reduction in hepatic blood flow
Decreased metabolic capacity
Genitourinary system
Kidney Decrease in glomerular filtration rate
Decrease in functioning renal glomeruli
Decrease in renal blood flow
Decrease in excretory and reabsorptive capacities of renal
tubules
Bladder Loss of supporting elastic tissue
Detrusor muscle hypertrophy
Reproductive tract Decrease in vaginal secretions
Increase in pH of vaginal secretions
 tv
~

Table 2. Drug interactions. (GOSNEY and TALLIS 1984)

Drug(s) Nature of clinically most important Comment No.

interaction cases
Frusemide Aminoglycosides Increased ototoxicity and nephrotoxicity Potentially serious, probably avoidable 24
Frusemide Cephalosporins Increased nephrotoxicity Potentially serious, probably avoidable 9
Warfarin Phenylbutazone Increased risk of haemorrhage Potentially dangerous, avoidable 1
Warfarin Aspirin Increased risk of haemorrhage Potentially dangerous, avoidable 3
Heparin Aspirin Increased risk of haemorrhage Potentially dangerous, avoidable 1
Verapamil Propranolol Risk of asystole Potentially dangerous, avoidability 2 ......
(within 6h) uncertain (1
~
Frusemide Prednisolone Antagonism Significant interaction, avoidability 16 n
(with potassium uncertain rn
t-'
supplementation) z
>
-<
Frusemide Prednisolone Increased risk of potassium loss as well Potentially serious, risk could have 14 po
~
(without potassium as antagonism (four of these had been reduced by potassium p..
supplementation) K+<2.8mmoVl) supplementation 'i:I
Loop diuretics Indomethacin Antagonises diuretic effect Potentially serious, could have been 4 ~
avoided by using alternative analgesic tI
Loop diuretics Other non-steroidal May antagonise diuretic effect Possibly significant, avoidability 15
;;
~
n
anti-inflammatory uncertain -<
Carbenoxolone Spironolactone drugs Inhibition of ulcer healing, interference Potentially significant, should have 1 ~
(JQ
with treatment of cardiac failure been avoided ('1)
I>'
Phenylbutazone Chlorpromazine Increased risk of leucopenia Potentially dangerous, avoidable 2 ::s
0..
Iron Antacid Interference with action/absorption Possibly significant, avoidable 2 CJ
('1)
::s('1)
Cimetidine Antacid Interference with action/absorption Possibly significant, avoidable 20 ....
(i'
Digoxin Antacid Interference with action/absorption Possibly significant, avoidable 5 'TI
I>'
(')
Tetracycline Antacid Interference with action/absorption Significant, avoidable 3 .....
0
...,
C/O
Tetracycline Iron Interference with action/absorption Significant, avoidable 2
S'
Warfarin Co-trimoxazole Interference with control, increased Significant, could have been avoided 5 0
...,
risk of bleeding by using trimethroprim alone ::s
(JQ

Phenytoin Co-trimoxazole Potentiation Possibly significant, could have been 1 ......

::s
avoided by using trimethoprim alone (D
...,
I>'
(')
Phenylbutazone Thyroxine Interference with thyroid function test Possibly significant, avoidable 1 .....
0'
Tolbutamide Co-trimoxazole Potentiation Possibly significant, could have been 2 ::sC/O
avoided by using trimethoprim alone
Total 133

tv
00
(;>
284 J.e. McELNAY and P.F. D' ARCY

I. Pharmacokinetics in the Elderly

There are striking examples of age-related alterations of both the kinetics and
dynamics of drugs. Every drug used in the elderly should have its pharmacol-
ogy and toxicology specifically evaluated in the elderly, yet few elderly pa-
tients, if any, participate in clinical trials on new drug substances. One does not
question that it is desirable, but it must be appreciated that the ethical situa-
tion is not at all clear. In a paper given to the Medico Pharmaceutical Forum,
GRIFFIN (1986a, cited in GRIFFIN 1989) stated that "on a personal basis I
believe that it is unethical to conduct studies in the elderly, where the elderly
person is being selected for the study by virtue of some presumed abnormality
of metabolism associated with the process of ageing. Such data should be
obtained in the course of clinical trials primarily to evaluate efficacy".
Pharmacokinetic effects of medication may be increased or decreased in
the elderly patient. POSNER and ROLAN (1994) have categorised age-related
differences in kinetics between the elderly and young as being primarily due
to: diminished renal function, altered proportions of body fat and water,
reduced cardiac output and some degree of altered hepatic metabolism. Al-
tered kinetics in the elderly are discussed below under the four main kinetic
processes of absorption, distribution, metabolism and elimination.

1. Absorption
The extent of drug absorption is generally little affected by age. It is true that
physiological changes occur with ageing that can influence drug absorption, for
example, decreased gastric acid secretion, decreased pancreatic lipase activity,
decreased gastric emptying and gastrointestinal motility, and decreased gastric
blood flow. However, although delays in absorption have been described for
some substances, they are generally of little clinical importance since the
extent of absorption remains largely unchanged. The absorption of levodopa
is an exception. It has been demonstrated, for example, that there can be a
threefold increase in its absorption in elderly patients probably due to an age-
related reduction in gastric dopamine decarboxylase and consequently greater
availability of levodopa for absorption into the systemic circulation (EVANS et
al. 1980). This increase in absorption is probably one of the main reasons for
the substantially lower dose of levodopa required for elderly patients (BROE
and CAIRD 1973) and for levodopa being one of the commoner causes of
adverse drug reactions (ADRs) in elderly patients.

2. Distribntion
Several blood flow changes occur with advancing years. Firstly there is a
decrease in cardiac output of 1 %/year after the age of 25 years. Secondly, in
older adults there is a decreased perfusion of limbs, liver and mesentery of up
to 45%. Thirdly, in elderly people there is a decreased blood flow to the heart
Age and Genetic Factors in Drug Interactions 285

(30%) and brain (15 %). Although little data are available, these changes in
blood flow could obviously influence the rate of drug distribution to various
tissues. In the elderly, however, it is the changes in body composition, includ-
ing the concentration of plasma albumin, which have the greatest impact on
drug distribution.
In the elderly, lean body mass is reduced, fat mass is increased and intra-
cellular fluid volume is decreased. The implications of these changes for drug
kinetics will depend largely on the degree of lipid and aqueous solubility of the
drug in question and its degree of binding to plasma albumin. The volume of
distribution of lipid-soluble drugs, e.g., diazepam and lignocaine, will obvi-
ously be increased while the opposite effect will be seen with water-soluble
drugs, e.g., gentamicin and nadolol. These effects will tend towards lower
plasma concentrations of lipid-soluble drugs and higher concentrations of
water-soluble drugs. Lipid-soluble drugs, as well as distributing into body fat,
can concentrate in the brain, which is high in lipoid tissue. There is also some
evidence to suggest that the blood-brain barrier is less intact in the elderly,
thus allowing drugs to distribute into the brain in increased concentrations.
Changes in drug distribution are of particular significance when the therapeu-
tic ratio of the drug is low as in the case of digoxin, where the 30% reduction
in its volume of distribution in the elderly is probably related to reduction in
lean body mass. This necessitates a reduction in the loading dose of digoxin in
elderly patients.
For drugs which are highly bound to plasma proteins, the decreased
concentrations of albumin that occur, particularly in sick or malnourished
elderly patients, can also have marked effects on drug distribution and there-
fore drug plasma levels. As a general rule, drugs which are highly bound to
plasma proteins distribute very little into organs since bound drug cannot cross
the capillary membranes, due to the large size of the protein molecules.
Warfarin is a good example of this effect - it is highly bound to plasma albumin
(>99%) and is largely retained in the plasma compartment. In the elderly,
however, since albumin concentrations are decreased, less drug is bound and
more drug is free to distribute throughout the body and to be eliminated by the
liver and kidney. This redistribution leads to decreased total concentrations
(free + bound) of drug in the plasma, but, in the absence of effects on drug
elimination, one would expect the free, pharmacologically active drug concen-
trations to remain relatively similar. In the elderly, however, drug elimination
is often decreased and therefore the free drug concentrations in plasma may
become elevated with consequent increases in pharmacological effects and
toxicity. When drugs are bound to albumin there is also a greater risk of
displacement due to interaction between drugs competing for the same bond-
ing sites if the serum albumin levels are reduced (see Chap. 4). In contrast to
albumin, plasma globulin concentrations tend to increase with increasing age.
This effect is relatively unimportant since few drugs are extensively bound to
these proteins.
286 J.e. McELNAY and P.F. D'ARCY

Some drugs are taken up actively into various tissues, particularly if they
resemble a naturally occurring nutrient. This method of tissue uptake is
utilised by some of the antitumour drugs, for example, the uptake of
melphalan, an amino acid derivative, into tumour cells has been shown to
involve active transport. One might expect the active uptake of drugs into
tissues to be decreased in the elderly due to declining metabolic function;
however, good data are not available to support this hypothesis.
In summary, the main effects of ageing on drug distribution include a
more widespread distribution of lipid-soluble drugs, more limited distribution
of water-soluble drugs and a more extensive distribution of drugs bound to
plasma albumin. A summary of the effects of ageing on drug distribution is
presented in Table 3 (RITSCHEL 1980).

Table 3. Change of drug parameters in the aged. (RnscHEL 1980)

Drug Biological Volume of Total

half-life distribution clearance

Acetaminophen (paracetamol) i U ,j,

Acetanilide i ,j, U
Aminopyrine i NI NI
Amitriptyline i ,j, ,j,
Amobarbital i U ,j,
Ampicillin i U ,j,
Antipyrine (phenazone) i ,j, ,j,
Aspirin NI i ,j,
Carbenicillin i i NI
Carbenoxolone i ,j, ,j,
Cefamandole i i u
Cefazoline i U ,j,
Cefradin i U ,j,
Chlordiazepoxide i i ,j,
Chlormethiazole i i ,j,
Chlorthalidone i U ,j,
Cimetidine i ,j, ,j,
Cyclophosphamide i i NI
Desipramine i NI NI
Desmethyldiazepam i i ,j,
Diazepam i i U or,j,
Digoxin i ,j, ,j,
Dihydrostreptomycin i NI NI
Doxycycline i U ,j,
Flurbiprofen U U U
Gentamicin i ,j, ,j,
Imipramine i NI NI
Indomethacin i NI NI
Indoprofen U U U
Kanamycin i NI ,j,
Levomepromazine i U ,j,
Lidocaine (lignocaine) i i U
Lithium NI NI ,j,
Lorazepam U or i ,j, U or,j,
Age and Genetic Factors in Drug Interactions 287

Table 3. Continued

Drug Biological Volume of Total

half-life distribution clearance

Methotrexate t J- J-
Metoprolol t NI NI
Morphine t NI NI
Netilmicin t U J-
Nitrazepam t U
Nortriptyline t U J-
Oxazepam t t NI
Penicillin G t NI NI
Phenobarbital t NI NI
Phenylbutazone t U J-
Phenytoin NI NI t
Procaine penicillin t NI NI
Practolol t J- NI
Propicillin t J- NI
Propranolol t J- J-
Protriptyline t J- J-
Quinidine t J- J-
Spironolactone t NI NI
Sulbenicillin t t NI
Sulfamethizole (sulphasomidine) t U or J- J-
Tetracylline t NI NI
Theophylline t U or J- J-
Thioridazine t NI NI
Tobramycin t J- J-
Tolbutamide t J- J-
Warfarin t U J-

3. Metabolism
Since the early animal experiments of CONNEY et al. (1956, 1957) and CONNEY
(1967) it has often been observed that the rate of oxidative metabolism of
various substrates can differ markedly depending on a number of factors of
which age is important. However, direct measurement of enzyme levels in liver
biopsy samples has shown few age-related alterations in the concentration of
several drug-metabolising enzymes in man (WOODHOUSE et al. 1984). Human
studies have suggested that the changes observed in clearance of oxidatively
metabolised drugs are more likely due to physiological changes in elderly
patients than to changes in their cytochrome P450 systems. There are few, if
any, detectable changes related to ageing in cytochrome P450 protein or
mRNA (SCHMUCKER and KONTAK 1990). Age-related reduction in overall liver
mass and absolute liver blood flow are most likely the important factors in
diminishing drug metabolism (CALLOWAY et al. 1965; SKAUNIC et al.1978). Age
may affect hepatic oxidative processes to a greater extent than conjugative
metabolic processes (see Table 4). Although an individual's liver function is
difficult to assess, problems are more likely in patients with ascites, jaundice
and encephalopathy.
288 J.e. McELNAY and P.F. D'ARCY

Table 4. Studies on the relation of age to the clearance of drugs by hepatic bio-
transformation. (GREENBLATI et al. 1982)

Drug or metabolite Initial pathway of biotransformation"

Evidence suggests age-related reduction

in clearance
Antipyrine b (phenazone) Oxidation (OR, DA)
Diazepamb Oxidation (DA)
Chlordiazepoxide Oxidation (DA)
Desmethyldiazepamb Oxidation (OR)
Desalkylflurazepamb Oxidation (OR)
Clobazamb Oxidation (DA)
Alprazolamb Oxidation (OR)
Quinidine Oxidation (OR)
Theophylline Oxidation
Propranolol Oxidation (OR)
Nortriptyline Oxidation (R)
Small or negligible age-related change
in clearance
Oxazepam Glucuronidation
Lorazepam Glucuronidation
Temazepam Glucuronidation
Warfarin Oxidation (OR)
Lidocaine (lignocaine) Oxidation (DA)
Nitrazepam Nitroreduction
Flunitrazepam Oxidation (DA), nitroreduction
Isoniazid Acetylation
Ethanol Oxidation (alcohol dehydrogenase)
Metroprolol Oxidation
Digitoxin Oxidation
Prazosin Oxidation
Data conflicting or not definitive
Meperidine (pethidine) Oxidation (DA)
Phenylbutazone Oxidation (OR)
Phenytoin Oxidation (OR)
Imipramine Oxidation (OR, DA)
Amitriptyline Oxidation (OR, DA)
Acetaminophen (paracetamol) Glucuronidation, sulfation
Amobarbital Oxidation (OR)

"OR denotes hydroxylation and DA dealkylation.

bEvidence suggests that the age-related reduction in clearance is greater in men than in
women.

4. Excretion
The two main routes by which drugs are eliminated from the body are metabo-
lism by liver enzymes, and excretion by the kidneys. The chemical changes
that result from liver metabolism generally (but not always) result in mol-
ecules that are less active pharmacologically and are less lipid soluble and/or
more easily ionised. This causes them to be more readily excreted by the
kidneys.
Age and Genetic Factors in Drug Interactions 289

In the geriatric patient, drug overdose is particularly likely to occur if the

drug remains active in the body until it is excreted by the kidneys. This is
because renal function (glomerular filtration and tubular excretion) diminish
with increasing age even in the absence of clinically detectable disease. A
reduction of about 30% in the glomerular filtration rate and tubular function
has been demonstrated in otherwise normal patients over 65 years of age when
compared with young adults. Indeed at 90 years of age the functional capacity
of the "normal" kidney may only be half what it was at 30 years of age (AGATE
1963). Diminished renal function due to age may be made worse by dehydra-
tion, congestive heart failure urinary retention and diabetic nephropathy, all
of which are more frequent in the elderly patient.
It has been suggested that the most important single kinetic cause of
adverse reactions to drugs in the elderly is impaired renal elimination (CAIRD
1985). This affects drugs eliminated by glomerular filtration (e.g., digoxin,
cimetidine and pro cain amide ) and those eliminated by tubular excretion (e.g.,
penicillins and aminoglycoside antibiotics). At all ages, the renal clearance of
digoxin is related linearly to creatinine clearance. The mean digoxin clearance
in elderly patients treated with the drug has been shown to be only 40mllmin
with values as low as lOmllmin as compared with the average normal glomeru-
lar filtration in age-comparable healthy elderly patients of 80-90mllmin
(ROBERTS and CAIRD 1976).

II. Pharmacodynamics in the Elderly

Pharmacodynamic interactions between drugs may produce similar or antago-

nistic pharmacological effects or side effects. They may be due to competition
at receptor sites, or occur between drugs acting on the same physiological
system. A large number of drug interactions are of no clinical significance and
it is really only interactions involving those drugs with a narrow therapeutic
index (e.g., lithium, anticoagulants, antiepileptics) and those which require
careful control of dosage (e.g., antihypertensives, antidiabetics) that are of
major clinical importance.
The pharmacodynamic effect of drugs can be influenced by age as illus-
trated in the following examples:
The elderly show an increased hypotensive response to angiotensin-con-
verting enzyme (ACE) inhibitors; they show a reduced responsiveness to
propranolol. Data are conflicting for the calcium antagonists. The inotropic
effect of theophylline is increased with age, but its bronchodilator effect is
decreased. The anticoagulant effect of warfarin is increased in elderly patients
due to the greater fragility of the hepatic synthesis of clotting factors
(SHEPHARD et al. 1977). Warfarin is a drug which is notorious for its involve-
ment in drug interactions and therefore this increased warfarin sensitivity
could predispose the elderly patient to greater adverse sequelae should an
interaction take place.
290 J.e. McELNAY and P.F. D' ARCY

FEELY and COAKLEY (1990) emphasised in their review On altered

pharmacodynamics in the elderly that the importance of age-related changes
in drugs sensitivity is increasingly appreciated. The type, intensity and dura-
tion of drug action can be affected, ranging from therapeutic failure to major
drug toxicity. Alterations in physiologic and homeostatic systems, including
the autonomic nervous system, baroreceptors, thermoregulation and balance,
have been described. The increased sensitivity of elderly patients to the pos-
tural hypotensive effects of medication has been well documented. These
effects can result in falls and dizzy spells (GRIBBIN et al. 1971; BAKER and
HARVEY 1985; TINETII et al. 1988), which can not only result in poor quality of
life in affected individuals but also can be catastrophic in relation to broken
bones, which often result from falls. RAMSAY (1981) showed that there was a
weak correlation between age and postural fall in diastolic pressure both
before and after treatment with methyldopa. He was careful, however, to
comment that this correlation might not be with age per se but with the
severity of the hypertension.
The elderly patient is less susceptible to developing tachycardia in re-
sponse to equivalent concentrations of isoprenaline (VESTA et al. 1979;
LAKATIA 1979). /3-Adrenoceptor antagonists are less effective at comparable
blood levels in elderly patients compared with the young (SCOTI and REID
1982); these effects are likely due to decreased numbers of available
/3-receptors.
Phenothiazines lower body temperature and are more likely to induce
hypothermia in the old than in the young. Similarly, elderly patients
are more likely to develop parkinsonian effects with phenothiazines than
younger patients (EXON-SMITH 1964; JONES and MEADE 1964; COLLINS et al.
1980).
Drugs affecting the CNS produce a relatively greater response in the
elderly than in the young for a given plasma concentration (WILLIAMS and
LOWENTHAL 1992). The elderly are particularly vulnerable to the side effects of
major tranquillisers. Antidepressant agents are particularly likely to cause
postural hypotension, urinary retention and sedation. Falls leading to hip
fractures have been associated with psychoactive drugs and antidepressants.
Nonsteroidal anti-inflammatory agents (NSAIDs) have an increased risk of
causing hyperkalaemia or renal failure or death from gastrointestinal
haemorrhage. They also cause sodium and fluid retention and so can reduce
the effects of drugs used to treat hypertension or congestive heart failure. This
latter pharmacodynamic interaction is often ignored.
With diuretics, the elderly are more susceptible to fluid and electrolyte
disorders, including volume depletion, hypokalaemia, hyponatraemia and
hypomagnesaemia. Adverse reactions to lignocaine are frequent in the elderly
(WILLIAMS and LOWENTHAL 1992). The increased effect of some benzodiaz-
epines is well known but it is uncertain whether this is due to receptor alter-
ations, or to the balance of neurotransmitters in the ageing brain, or perhaps
to better penetration of the blood-brain barrier.
Age and Genetic Factors in Drug Interactions 291

All of the drugs mentioned in this (pharmacodynamics) and the preceding

section (pharmacokinetics) are commonly susceptible to interaction with
other drugs, and if interactions occur, and if they potentiate the activity of one
of the component drugs, then they will almost certainly present a greater
hazard than if they occurred in a younger age group. The results of the
interaction will be qualitatively similar in the different age groups but more
pronounced in its effects in the elderly.

III. Inappropriate, Unnecessary and Interacting Medication

In a paper entitled "Inappropriate Prescribing in the Elderly", ADAMS et al.
(1987) investigated the medication of 1094 patients who were pensioners
attending the Accident and Emergency Department of a United Kingdom
District General Hospital. Of these 871 (79.6%) were taking at least one
prescribed medication. The most commonly consumed medications as a per-
centage of all subjects in the study were diuretics 32%, analgesics and non-
steroidal anti-inflammatory agents 25.2%, hypnotics and sedatives 18.5%,
bronchodilators 14.5% and nitrites 10%. Of all the prescriptions written, 4.8%
were identified as having been written for conditions for which they were
contraindicated according to the British National Formulary. In addition, 356
interacting combinations were identified in the 2353 prescriptions, affecting
216 (19.7%) of patients taking medication. Therefore in this study at least
24.5% of prescriptions written for elderly patients were inappropriate.
LINDLEY et al. (1992) carried out a similar study. They examined 416
successive admissions of elderly patients to a United Kingdom teaching hospi-
tal. Interacting drug combinations and drugs with relative contraindications
were common, but not as important in producing ADRs as drugs with absolute
contraindications or unnecessary drugs. Forty-eight patients (11.5% of admis-
sions) were taking a total of 51 drugs with absolute contraindications (3.8% of
prescriptions ).
One hundred and seventy-five drugs were discontinued on or shortly after
admission in 113 (27%) patients because they were deemed to be unnecessary.
One hundred and three patients (27%) of those on medication experienced
151 ADRs of which 75 (49.7%) were due to drugs with absolute con-
traindications and/or that were unnecessary, a significantly higher rate of
ADRs than observed for all prescriptions. Of 26 (6.3%) admissions attributed
to ADRs, 13 (50%) were due to inappropriate prescriptions. The admission
rate per prescription was significantly higher for inappropriate than for appro-
priate drugs. The authors concluded that much drug-related morbidity in the
elderly population was due to inappropriate prescribing.
Some idea of the relative frequency of drug interactions among elderly
patients has been given by more specific publications:
GOSNEY and TALLIS (1984) from Liverpool, United Kingdom, surveyed,
retrospectively, the prescription of contraindicated and interacting drugs in
elderly patients admitted to hospital. Contraindicated or adversely interacting
292 J.e. McELNAY and P.F. D' ARCY

drugs were identified in 200 (3.2%) of 6160 prescriptions. One hundred and
thirty-six (23.7%) patients were affected. The most common interactions (po-
tential or actual) were frusemide with aminoglycosides (increased ototoxicity);
cimetidine with antacids (interference with action/absorption); frusemide with
prednisolone (potassium loss); loop diuretics with NSAIDs (antagonism
of diuretic effect), and frusemide with cephalosporins (increased
nephrotoxicity).
NOLAN and O'MALLEY (1989) evaluated the risk of potential drug inter-
actions in 11 private nursing homes in Dublin. The percentage of patients
prescribed potentially interacting combinations increased greatly with the
number of medications taken.
Potential drug-drug interactions were also studied in an ambulatory popu-
lation in Florida (HALE et al. 1989). Ten major interaction classifications were
studied. It was found that 40% of patients taking quinidine were also taking a
digitalis glycoside, and nearly one-third of patients taking warfarin were also
taking a drug with an interacting potential.
SCHNEIDER et al. (1992) investigated adverse drug reactions in an elderly
outpatient population attending an interdisciplinary geriatric clinic and a
medical clinic in Cleveland, Ohio, United States. The sample size of the study
was 463 patients, of whom 332 attended the medical clinic and 131 attended
the geriatric clinic. Potential drug interactions were identified in the records of
143 (31 %) sUbjects.
DOUCET et al. (1993), from Rouen and Bois Guillaume, studied drug
interactions in French patients over 65 years old. Of 513 elderly patients
admitted to hospital, the principal drug interactions occurred with diuretics
and benzodiazepines. Of this population 63 % had one or more drug interac-
tions leading to an adverse effect (124 patients) and to hospitalisation in half
these cases.
STANTON et al. (1994) studied drug-related admissions to an Australian
hospital. The median age of the patients was 67 years; 4.4 % of admissions were
due to drug interactions. SCHENKER and BAY (1994) working in a Veterans
Medical Administration Center in Texas, United States, investigated drug
disposition and hepatotoxicity in the elderly and concluded that drug-drug
interactions and concurrent derangements accompanying advanced age made
a significant contribution to adverse drug effects.
Apart from these group studies, there have been a large number of single
patient reports in which elderly patients suffered drug interactions. For ex-
ample, KERR (1993) in the United States described an elderly woman who
developed gastrointestinal bleeding probably potentiated by the interaction of
fluconazole with warfarin. SCARFE and ISRAEL (1994) also in the United States,
reported the case of an elderly woman who experienced a significant increase
in INR after levamisole and fluorouracil were added to an established regimen
of warfarin.
These studies confirm the obvious, that as the number of medications in a
patient's regimen increases, the potential for interacting combinations also
Age and Genetic Factors in Drug Interactions 293

increases. Multiple drug therapy and differences in pharmacokinetics and

pharmacodynamics are thus factors that predispose the elderly to adverse
reactions to drugs, e.g., to digoxin, to diuretics and to psychotropic agents
(NOLAN and O'MALLEY 1989). Therefore, if elderly patients also receive inap-
propriate medication (commonly digoxin, diuretics and psychotropic agents),
or interacting combinations, and if they are subject to age-related alterations
of drug dynamics and kinetics, then it is quite obvious that drug-drug interac-
tions that potentiate the activity of one of the components might well present
a particular and peculiar hazard.

IV. Logistics: Age and Medicine Consumption

A World Health Organisation report on health care in the elderly (WORLD
HEALTH ORGANIZATION 1981) concluded that elderly patients consume a lot of
drugs, that polypharmacy seems to be the rule in acute hospital settings and in
institutions, and that psychoactive drugs appear to be used where most of the
mentally disabled are found. One may question whether the position has
changed in the 14 years that has elapsed since this report was published.
Throughout the past 40 years, Sweden, Norway and the United Kingdom
have recorded the highest ratio of elderly people (men aged over 65 years,
women over 60 years) to population among the industrialised countries
(Organisation for Economic Cooperation and Development, OECD). The
projected figures for the percentage changes in United Kingdom elderly popu-
lation for the years 1990-2001 (and 2001-2011) are 2.4% (5.0%) over the age
of 65 years, 11.0% (1.5%) over the age of 75 years, and 34.6% (12.8%) over
the age of 85 years. United Nations has calculated for OECD countries that
the figure for elderly persons (aged 65 and over) expressed as a percentage of
the total population will be at the year 2000: 15.2% for the United Kingdom,
12.8% for the United States and an average of 13.9% for the OECD countries
(OFFICE OF HEALTH ECONOMICS 1992).
Medication prescribed for the elderly in the National Health Service rose
by 29% from 98 million items in 1978 to 143 million by 1988. The total number
of prescription items was 350 million in 1988 for the whole population; thus the
elderly'S share of the total advanced from 32% to 41 % during the same period.
On average the elderly had between them 16.3 prescription items per annum
in 1988, as compared with 12.2 items in 1978. In sharp contrast, the number of
prescriptions given to those of working age fell by 10% from 6.1 items per
person to 5.5 items between 1978 and 1988 (OFFICE OF HEALTH ECONOMICS
1989). In the United States, it seems that geriatric patients consume 33% of
prescription drugs (SLOAN 1992).

V. Long-Term Treatments
Most of the drugs involved in clinically relevant interactions are those on
which patients are carefully stabilised for relatively long periods (e.g., oral
294 J.e. McELNAY and P.F. D'ARCY

anticoagulants, anticonvulsants, oral antidiabetics, cardiovascular drugs, psy-

chotropic agents).
Past experience has clearly shown that the elderly often receive these
types of medications, and it is the elderly, drug-stabilised patient who will be
at special risk of any changes in therapy or environment that will influence the
potency or bioavailability of normal medication. It is under these circum-
stances that drug-drug interactions may become dangerous to the elderly
patient (see, for example, Table 2). In order to make drug therapy safer
in elderly patients, two main strategies may be employed: Firstly, improve
patient compliance with medication instructions. Secondly, improve patient
prescribing and monitoring via avoiding unnecessary treatments, reviewing
drug therapy regularly, simplifying drug regimens and monitoring for adverse
reactions and interactions.

c. Genetic Factors
The term "pharmacogenetic disorder" was originally coined by VOGEL (1959)
and was originally limited to hereditary disorders revealed solely by the use of
drugs. Its meaning has now been enlarged to embrace all genetic contributions
to the considerable variation that exists in the interaction between man and
the pharmacological agents that he uses. The term can therefore be taken to
cover the adverse drug reactions and interactions that occur when particular
drugs are given to a patient with a genetically determined dysfunction of an
organ or body system, for example, renal polycystic disease.
As with the effect of age on the incidence or severity of drug-drug interac-
tions, there is no reason to suggest that genetic influences cause drug interac-
tions per se. What is clear, however, is that patients who suffer the effects of
pharmacogenetic disorders may well show a predisposition to specific drug-
drug interactions, or may show enhanced effects when such interactions occur.
The risk of drug toxicity usually arises from enzyme deficiency states.
Examples of these have been well described by BENNETI (1993). For example:
Hepatic porphyrias, where deficient conversion of porphyrins to haem
exposes affected individuals to risk from drugs which have the common prop-
erty of increasing the activity of delta-aminolaevulinic acid synthetase, the
rate-limiting enzyme of porphyrin synthesis, in their livers. Disorders of por-
phyrin metabolism have been comprehensively reviewed by FLETCHER and
GRIFFIN (1986) and they have presented a figure of the overall scheme of
porphyrin synthesis and a list of those drugs which have been recognised as
being associated as precipitating agents of porphyria. The figure (Fig. 2) and
list of drugs (Table 5) are reprinted here.
Malignant hyperpyrexia is a rare pharmacogenetic disease (rigid and non-
rigid types) occurring both in man and in the Landrace strain of pigs. In
susceptible individuals any potent inhalation anaesthetic or any skeletal
muscle relaxant can cause fever, rigidity, hyperventilation, cyanosis, hypoxia,
 'INHERITED DISORDER
Biochemical detect in porphyria believed
Acetylating mechanisms to be an excessive activity of ALA synthetase,
(Acetylchotine) the rate controlling enzyme of the synthesis.
Various forms of disease probably due to
~
(l)
disturbances lower down chain resulting
ACEm.COA< from overactivity of ALA synthelase. §
0..
Other melabolic pathways /
C'l
(l)
presence of ~
(l)
1
SUCCINYL co A. + GLYCINE -----.. a AMINO-f}OXO ADIPIC ACID O.
(')
pyridoxal
phosphate loss of C02 under influence of ALA synthetase
1 "TI
~
(')
1 condensation 8'
PORPHOBILINOGEN (PBG) I>-AMINO LAEVULINIC ACID (ALA) ....
of two molecules '"5'
under influence
of aminolaevutate o
dehydratase
Four molecules of 1 &
PBG under influence INTERMEDIATE PRODUCTS ......
~
of porphobilinogenase Uroporphyrinogen I Uroporphyrin I /urine 100)1g porphyrin/day rl
• (l)
combine 0.7mg ....
formed (TYPE 1) TYPE 1 ~

~ perday ~
Coproporphyrinogen I Coproporphyrin I o·
• \ I '\. ,,,'~ "'"'" ".",rio"" ~
1 '"
dehydrogenation EXCRETED
UROPORPHYRINOGEN III Uroporphyrin III Urine 50)1g porphyrin/day

decarboxylation 300 mg/day formed

• (TYPE III)
I \ /
t TYPE III "...
dehydrogenation
COPROPORPHYRINOGEN III Coproporphyrin III Faeces 150)1g porphyrin/day
• \

1 decarboxylation 'DRUG ACTION

enzyme ferrochelalase Excretion of type III porphyrins
in urine and faeces increases
PROTOPORPHYRINOGEN IX Haem ~ Fe++ in diseases of liver and in
haemolytic anaemia.
t dehydrogenation t..-protein

PROTOPORPHYRIN r Haemoproteins t
Haemogiobin ~ Cytochromes
Myoglobin
t5
VI

Fig. 2. Overall scheme of porphyrin synthesis. (FLETCHER and GRIFFIN 1986)

296 J.e. McELNAY and P.F. D'ARCY

Table 5. Drugs classified as to their association with porphyria. (FLETCHER and GRIFFIN
1986)

Potentially Drugs believed not Potentially Drugs believed not

porphyrogenic drugs to precipitate porphyrogenic drugs to precipitate
porphyria porphyria

Alcohol Acetazolamide Flufenamic acid Heparin

Alphaxalone Adrenaline Flunitrazepam Hydrallazine
Aluminium Alclofenac Fluroxine Hyoscine
2-Allyloxy-3 Amitriptyline Frusemide
methylbenzamide Aspirin Ibuprofen
Amidopyrine (A) Atropine Glutethimide Imipramine
Aminoglutethemide Gold preparations Indomethacin
Amitriptyline B vitamins (except Griseofulvin (A) Insulin
Amphetamines pyridoxine)
Androgens (A, C) Bethanidine Halothane Ketamine
Apronalide Biguanides Hydantoins Ketoprofen
A vapyrazone Bromides (phenytoin,
Azapropazone Bumetanide ethotoin, Labetolol
Bupivacaine mephentoin) Lithium
Barbiturates (A, C) Buprenorphine Hydrallazine Lorazepam
Bemegride Hydrochlorthiazide
Busulphan Cephalexin Hyoscine N-butyl Mandelamine
Cephalosporins bromide Mecamylamine
Carbromal Chloral hydrate Meclozine
Carbamazepine Chloramphenicol Imipramine Mefenamic acid
Chlorambucil Chlormethiazole Indomethacin Mersalyl
Chloramphenicol Chlorpheniramine Metformin
Chlordiazepoxide Chlorpromazine Isoniazid Methadone
(A) Chloroquine Isopropylmeprobamate Methenamine
Chlormezanone Chlorthiazides Lignocaine mandelate
Chloroform Clobazam Methylphenidate
Chloroquine (C) Clofibrate Mefenamic acid Morphine
Chlormethiazole Clonazepam Mephenazine
Chlorpropamide Codeine Meprobamate Naproxen
(A, C) Colchicine Mercury compounds Neostigmine
Cimetidine Corticosteroids Methoxyflurane Nitrofurantoin
Clonazepam Cyclizine Methsuximide Nitrous oxide
Clonidine Cyclopropane Methyldopa Nortriptyline
Cocaine Methyprylone
Colistin Dexamethasone Metoclopramide Oxazepam
Cyclophosphamide Diamorphine Metyrapone Oxpentifylline
Diazepam Metronidazole
Danazol Diazoxide Paracetamol
Dapsone Digitalis compounds Nalidixic acid Paraldehyde
Diazepam Diphenhydramine Nikethamide Penicillamine
Dichloralphenazone Dicoumarol Novobiocin Penicillins
(A) anticoagulants Nitrazepam Pethidine
Diclofenac Disopyramide Nitrofurantoin Phenformin
Diethyhexyl Domperidone Pheniramine
phthalate Droperidol Oral contraceptives Phenothiazines
Diethylpropion Oxazolidinediones (e.g.,
Dimenhydrinate EDTA (paramethadione chlorpromazine)
Ether (diethyl) and trimethadione) Phenoperidine
Enflurane Oestrogens (A, C) Phenylbutazone
Ergot preparations Fentanyl Oxazepam Prednisolone
Erythromycin Flurbiprofen Procaine
Ethchlorvynol Fusidic acid Pancuronium Prilocaine
Ethinamate Paraldehyde Promethazine
Ethosuximide Gentamicin Pargyline Propantheline
Etomidate Glipizide Pentazocine bromide
Eucalyptol Guanethidine Pentylenetetrazol Propanidid
Age and Genetic Factors in Drug Interactions 297

Table 5. Continued
Potentially Drugs believed not Potentially Drugs believed not
porphyrogenic drugs to precipitate porphyrogenic drugs to precipitate
porphyria porphyria
Pethidine Primaquine Spironolactone Tubocurarine
Phenazone Propoxyphene Steroids Sodium valproate
Phenelzine Propranolol Streptomycin
Phenoxybenzamine Prostigmine Succinimides Vitamin C
Phensuximide Pyrimethamine (ethosuximide,
Phenylbutazone methsuximide,
Phenylhydrazine Quinine phensuximide)
Primidone Sulphonal
Probenecid Reserpine Sulphonamides (A)
Progestogens (A, C) Resorcinol Sulphonylureas
Propanidid Rifampicin Sulthiame
Pyrazinamide
Pyrazolones Streptomycin Tetracyclines
(antipyrine, Succinylcholine Theophylline
isopropylantipyrine, Tolazamide
dipyrone, sodium Tetracyclines Tolbutamide (A, C)
phenyl dimethyl Tetraethylammonium Tranylcypromine
pyrazolone) bromide Trional
Pyridoxine Thiouracils Troxidone
Pyrimethamine Tricyclic
antidepressants Valproic acid
Ranitidine (amitriptyline) Viloxazine
Rifampicin Trifluoperazine
Thiazides Xylocaine
Sodium valproate Tripelennamine

A, acute porphyria; C, cutaneous porphyria.

Substances marked in italics have been variously described as causing porphyria or being safe to
administer to patients with porphyria.

respiratory and metabolic acidosis, and hyperphosphataemia with a raised

glucose level. Initial hyperkalaemia and hypercalcaemia are followed by
hypokalaemia and hypocalcaemia. The mortality of the condition is high
(60%-70%) (BRITI and KALOW 1968; BRITI et al. 1969; FURNISS 1970; ISAACS
and BARLOW 1970; DENBOROUGH et al. 1970, 1973; CARPOPRESSO et al. 1975;
SUTHERLAND and CARTER 1975).
Glucose-6-phosphate Dehydrogenase Deficiency. The life-span of the erythro-
cyte in a normal subject varies between 110 and 125 days. In haemolytic
anaemias the red cells are destroyed more rapidly and the average life-span of
the erythrocyte is correspondingly shorter. It has been known for some years
that, when certain ordinary, otherwise harmless oxidant drugs are adminis-
tered to some individuals in ordinary therapeutic doses, this is accompanied by
rapid red-cell destruction and the development of acute haemolytic anaemia
(Table 6). The cause of this can be an inherited deficiency of glucose-6-
phosphate dehydrogenase (G6PD). GRIFFIN (1986b) has reviewed the types of
G6PD deficiency [e.g., African type (A), Mediterranean type, the variants
Canton and Union] and has also discussed the occurrence of other types of
haemolytic anaemias not limited to patients with G6PD deficiency.
298 J.e. McELNAY and P.F. D'ARCY

Table 6. Adverse reactions due to poor microsomal

oxidation. (SMITH and SHAH 1984, personal commu-
nication; SMITH and IDLE 1981)

Adverse reaction Drugs concerned

Lactic acidosis Metformin

Agranulocytosis Carbimazole
Phenylbutazone
Chlorpromazine
Nortriptyline
Imipramine
Thioridazine
Captopril
Neuropathy or sensory Phenytoin
disturbances
Hepatic adenoma Oral contraceptives
Cerebellar signs Phenytoin
Vitamin-D-deficiency-like state Phenytoin
Phenobarbitone
Folate-deficiency-like state Phenytoin
Phenobarbitone
Syncope Prazosin
Malignant ventricular arrhythmias Mexiletine
Disopyramide
Bradycardia Propranolol
Metoprolol

Defective Carbon Oxidation (Oxidative Genetic Polymorphism). Mono-oxi-

dation represents the single most important step in hepatic metabolism of
drugs taking place in the P450 system. Heterogeneity of this oxidative system
has been well described (see Chap. 5). The first to be documented related to
the genetic polymorphism of the hydroxylation of the antihypertensive agent
debrisoquine and this has been used extensively as a tool to further elucidate
the nature of genetic polymorphism.
Affected individuals exhibit adverse responses to standard doses of drugs
whose inactivation involves oxidation of their carbon centres. In general 90%
of Caucasian populations are extensive metabolisers and 10% poor
metabolisers. The latter are virtually unable to oxidise and so the unchanged
drug is liable to accumulate with adverse effect. Table 6 summarises the
published work of SMITH and IDLE (1981) and the communicated results of
SMITH and SHAH (1984, personal communication). The table shows those ad-
verse drug reactions that are considered to have a metabolic basis and that are
attributable to interindividual differences in oxidative status. BENNETT (1993)
has added additional information: increased j3-blockade with bufuralol and
Age and Genetic Factors in Drug Interactions 299

timolol, postural hypotension with nortriptyline, and prolonged cardiovascu-

lar action with nifedipine.
Acetylator Status. Many drugs are acetylated during their metabolism, for
example, the hypotensive hydralazine, the antitubercular isoniazid, the
monoamine oxidase inhibitor (MAOI) phenelzine, and the sulphonamides.
Subjects vary in their speed of acetylation. This variation is genetically
linked and this acetylation polymorphism is an excellent example of a
pharmacogenetic phenomenon that is clinically relevant. Slow acetylators
respond adversely to standard doses of hydralazine and procainamide
(antinuclear antibodies in plasma, and some proceed to systemic lupus
erythematosus) and dapsone (haemolytic anaemia). On the other hand, fast
acetylators often respond less favourably to treatment because they have
lower blood levels of a drug and suffer from inadequate dosage. They also
appear to be at a greater risk from treatment with isoniazid, which may cause
acute hepatocellular necrosis due to the formation of an active metabolite
(BENNETT 1993).

Abnormalities of Plasma Pseudocholinesterase. Variation in the duration of

the neuromuscular blocking action of suxamethonium is probably the classical
example of this particular category. Suxamethonium is a useful and potent
neuromuscular blocking agent of the depolarising type with a very brief dura-

Table 7. Drugs reported to induce haemolysis in subjects with G6PD deficiency

Drug Haemolysis

Black subjects Caucasian subjects

Acetanilide +++++
Dapsone ++ +++
Furazolidone ++
Nitrofural ++++
Nitrofurantoin ++ ++
Sulphanilamide +++
Sulphapyridine +++ +++
Sulphacetamide ++
Salazosulphapyridine +++
Sulphamethoxypyridazine ++
Thiazosulphone ++
Quinidine ++
Primaquine +++ +++
Pamaquine ++++
Pentaquine +++
Quinocide +++ ++
Naphthalene +++ +++
Neoarsphenamine ++
Phenylhydrazine +++
Toluidine blue ++++
Trinitrotoluene +++
300 J.e. McELNAY and P.F. D' ARCY

tion of effect. This is because suxamethonium is rapidly hydrolysed and inac-

tivated by pseudocholinesterase in the serum.
There have been reports, since suxmethonium was introduced in 1949 by
BOVET et aI., of prolonged apnoea and muscular pain and stiffness in some
patients (BOURNE et aL 1952; EVANS et aL 1952; LEHMANN and RYAN 1956;
KAUFMAN et aL 1960; TELFER et aL 1964). Investigation showed that these
adverse effects were associated with low plasma pseudocholinesterase levels.
A familial incidence of abnormal sensitivity to suxamethonium was demon-
strated in patients who were shown to have an atypical pseudocholinesterase
activity due to genetic abnormalities. Four allelic genes seem to control the
inheritance of pseudocholinesterase; one normal, two atypical and one silent
allelic gene. They form pseudocholinesterases with varying activity. Such pa-
tients do not have any other recognised abnormality and they usually present
as cases of prolonged apnoea after a single injection of suxamethonium.
This atypical enzyme hydrolyses suxamethonium at a much slower rate
than does the normal type of serum cholinesterase. Consequently, the apnoea
that the drug induces is excessively prolonged. Fortunately, KALOW et aL
(KALOW and GENEST 1957; KALOW and STARON 1957) were able to develop a
technique to detect the presence of this atypical enzyme.
Raised Intraocular Pressure and Glaucoma with Topical Glucocorticoids.
Repeated topical application of glucocorticoids to the eye is followed by an
increase in intraocular pressure (ARMALY 1968; SCHWARTZ et aL 1972). The
extent to which this happens depends on the age of the subject and his/her
genetic make-up. The European population studied showed a trimodel
distribution: 66% with low, 29% with intermediate, and 5% with high-pressure
changes in response to the local application. Familial studies have suggested
that the response is influenced by two alleles pL and pH, with the three groups
represented by the genotypes pLpL, pLpH, and pHpH, respectively. The rise in
intraocular pressure is totally reversible by withdrawal of the drug.
The risk of developing open-angle hypertensive glaucoma spontaneously
is greatly increased in patients with pHpH genetic make-up. Although the other
genotypes are found in cases of glaucoma, the pLpL is much smaller than could
be expected for the normal population.

D. Comment
This chapter is not able to show that age or genetic make-up is directly
responsible for drug-drug interactions. The examples of therapeutic hazard
that have been given in this text are largely those due to adverse drug reac-
tions. It has been shown, however, that both age and genetic factors can
seriously affect drug therapy either by reducing its effect or alternatively
enhancing its effect, including increased toxicity.
It has been shown in many studies that the elderly as a group have to take
more medicines than younger age groups because of the disease conditions
Age and Genetic Factors in Drug Interactions 301

that occur as age progresses. It is also a well-established fact that the more
medicines that are taken together the greater the possibility of drug-drug
interactions occurring. It therefore follows that the elderly are likely to suffer
the effects of more drug-drug interactions than younger age groups and that,
due to decreased resilience in the elderly, such interactions will probably have
greater effect on their continued health than similar interactions would have
on younger patients.
Genetic abnormalities can increase the toxicity of drug therapy at normal
therapeutic dosage levels. This may therefore increase the possibility of hazard
from specific drug-drug interactions. Medical opinion does not generally
favour multiple-drug therapy, but on occasion it is necessary. The price that
may be required for using multiple drugs in elderly patients and in those
persons with an abnormal genetic make-up may be the occurrence of relatively
more and more serious drug-drug interactions.

References
Adams KRH, Al-Hamouz S, Edmund E, Tallis RC, Vellodi C, Lye M (1987) Inappro-
priate prescribing in the elderly. J R Coll Physicians Lond 21:39-41
Agate J (1963) The practice of geriatrics. Heinemann, London, pp 217-222
Armaly MF (1968) Genetic factors related to glaucoma. Ann NY Acad Sci 151:861-
875
Baker SP, Harvey AH (1985) Fall injuries in the elderly. Clin Geriatr Med 1:501-512
Bennett PN (1993) Evaluation of toxicity in human subjects. In: Ballantyne B, Marrs T,
Turner P (eds) General and applied toxicology, voll. Macmillan, Basingstoke, pp
367-384
Bourne JG, Collier HOJ, Somers GF (1952) Succinylcholine (succinoylcholine)
muscle-relaxant of short action. Lancet 1:1225-1229
Bovet D, Depierre F, Courvoisier S, de Lestrange Y (1949) Synthetic curarizing agents.
II. Phenolic ethers with quaternary ammonium groups. The action of tris
(diethylamino-ethoxy) benzene triiodoethylate (2559F). Arch Int Pharmacodyn
80:172-188
Braverman AM (1982) Therapeutic considerations in prescribing for elderly patients.
Eur J Rheumatol Infiamm 5:298-293
Britt BA, Kalow W (1968) Hyperrigidity and hyperthermia associated with
anaesthesia. Ann NY Acad Sci 151:947-958
Britt BA, Locher WG, Kalow W (1969) Hereditary aspects of malignant hyperthermia.
Can Anaesth Soc J 16:89-98
Broe GA, Caird PI (1973) Levodopa for parkinsonism in elderly and demented pa-
tients. Med J Aust 1:630--635
Caird FI (1985) Towards rational therapy in old age. The F.E. Williams Lecture 1985.
J Roy Coll Physicians Lond 19:235-239
Calloway NO, Foley CF, Lagerbloom P (1965) Uncertainties in geriatric data. II.
Organ size. J Am Geriatr Soc 13:20-28
Carbonin P, Pahor M, Bernabei R et al. (1991) Is age an independent risk factor of
adverse drug reactions in hospitalized medical patients? J Am Geriatr Soc
39:1093-1099
Carpopresso PR, Gittleman MA, Reilly DJ, Paterson LJ (1975) Malignant
hyperthermia associated with enfiurane anaesthesia. Arch Surg 110:1491-1493
Chrischilles EA, Segar ET, Wallace RB (1992) Self-reported adverse drug reactions
and related resource use. A study of community-dwelling persons of 65 years of
age and older. Ann Intern Med 117:634-640
302 J.e. McELNAY and P.F. D' ARCY

Collier P, Maguire T, Morrow N, McCollum A, McElnay J, Scott E, Scott M, Singleton

M (1993) Care for the elderly. A self-study course for pharmacists. Northern
Ireland Centre for Postgraduate Pharmaceutical Education and Training of Phar-
macists, and the Pharmacy Practice Group, School of Pharmacy, The Queen's
University of Belfast
Collins KJ, Exton-Smith AN, James MH, Oliver JD (1980) Functional changes in
automatic nervous responses with ageing. Age Ageing 9:17-24
Conney AH (1967) Pharmacological implications of microsomal enzyme induction.
Pharmacol Rev 19:317-366
Conney AH, Miller EC, Miller JA (1956) The metabolism of methylated aminoazo
dyes. Evidence for the induction of enzyme synthesis in the rat by 3-methylcholan-
threne. Cancer Res 16:450-459
Conney AH, Miller EC, Miller JA (1957) Substrate-induced synthesis and other prop-
erties of benzpyrene hydroxylase in rat liver. J BioI Chern 228:703-706
D'Arcy PF, McElnay JC (1983) Adverse drug reactions and the elderly patient. Adv
Drug React Ac Pois Rev 2:67-101
Denborough MA, Ebeling P, Kimng JO, Zapf P (1970) Myopathy and malignant
pyrexia. Lancet 1:1138-1140
Denborough MA, Dennett X, Anderson RMcD (1973) Central core disease and malig-
nant hyperpyrexia. Br Med J 1:272-273
Denham MJ (1990) Adverse drug reactions. Brit Med Bull 46:53-62
Doucet J, Chassagne P, Pauty MD, Breton T, Drieux S, Hemet C, Denis P, Menard JF,
Bercoff E (1993) Drug-interactions in patients over 65 - effects and their types -
prospective-study of 517 cases. Rev Med Interne 14:439
Evans FT, Gray PWS, Lehmann H, Silk E (1952) Sensitivity to succinylcholine in
relation to serum cholinesterase. Lancet 1:1229-1230
Evans MA, Triggs EJ, Broe GA, Saines N (1980) Systemic activity of orally adminis-
tered L-dopa in the elderly Parkinsonian patient. Eur J Clin PharmacoI17:215-222
Exon-Smith AN (1964) Accidental hypothermia in the elderly. Br Med J 2:1255-1258
Feely J, Coakley D (1990) Altered pharmacodynamics in the elderly. Clin Geriatr Med
6:269-283
Fletcher AP, Griffin JP (1986) Disorders of porphyrin metabolism. In: D'Arcy PF,
Griffin JP (eds) Iatrogenic disease, 3rd edn. Oxford Medical, Oxford, pp 67-81
Furniss P (1970) Hyperpyrexia during anaesthesia. Br Med J 4:745
Gosney M, Tallis R (1984) Prescription of contraindicated and interacting drugs in
elderly patients admitted to hospital. Lancet 11:564--567
Greenblatt DJ, Sellers EM, Shader RI (1982) Drug disposition in old age. N Engl J
Med 306:1081-1088
Gribbin B, Pickering TG, Sleight P, Peto R (1971) Effect of age and high blood
pressure on baroreflex sensitivity in man. Circulation 29:424--431
Griffin JP (1986a) Lecture given to the Medico Pharmaceutical Forum, October 1986,
cited in Griffin JP (1989) Drugs and the elderly. In: Griffin JP (ed) Medicines:
regulation research and risk. Greystone Books, The Queen's University of Belfast,
p203
Griffin JP (1986b) Pharmacogenetic and iatrogenic disease. In: D' Arcy PF, Griffin JP
(eds) Iatrogenic diseases, 3rd edn. Oxford Medical, Oxford, pp 59-66
Griffin JP (1992) Drugs and the elderly. In: Griffin JP (ed) Medicines: regulation
research and risk. Greystone Books, The Queen's University of Belfast, pp 373-
387
Gurwitz JH, Avorn J (1991) The ambiguous relation between aging and adverse drug
reactions. Ann Intern Med 114:956-966
Hale WE, Monks RG, Stewart RB (1989) Drug use in a geriatric population. J Am
Geriatr Soc 27:374--377
Isaacs H, Barlow MB (1970) Malignant hyperpyrexia during anaesthesia: possible
association with subclinical myopathy. Br Med J 1:275-277
Jones IH, Meade TW (1964) Hypothermia following chlorpromazine therapy in
myxoedematous patients. Gerontol Clin (Basel) 6:252-256
Age and Genetic Factors in Drug Interactions 303

Kalow W (1956) Familial incidence of low pseudocholinesterase level. Lancet 11:576-

577
Kalow W (1959) In: Wolstenholme GE, O'Conner MC (eds) Ciba Foundation Sympo-
sium on the biochemistry of human genetics. Churchill-Livingstone, London, p 39
Kalow W, Genest K (1957) A method for the detection of atypical forms of human
serum cholinesterase. Determination of dibucaine number. Can J Biochem
PhysioI35:339-346
Kalow W, Staron N (1957) On distribution and inheritance of atypical forms of human
serum cholinesterase, as indicated by dibucaine numbers. Can J Biochem Physiol
35:1305-1320
Kaufman L, Lehmann H, Silk E (1960) Suxamethonium apnoea in an infant: expres-
sion of familial pseudocholinesterase deficiency in three generations. Br Med J
1:166-167
Kellaway GSM, McCrae E (1973) Intensive monitoring for adverse drug effects in
patients discharged from acute medical wards. N Z Med J 78:525-528
Kerr HD (1993) Potentiation of warfarin by fluconazole. Am J Med Sci 305:164-165
Lakatta EG (1979) Alterations in the cardiovascular system that occur in advanced age.
Fed Proc 38:163-167
Lamy PP (1991) Physiological changes due to age. Pharmacodynamic changes of drug
action and implications for therapy. Drugs Aging 1:385-404
Landahl S (1987) Drug treatment in 70-82-year-old persons. A longitudinal study. Acta
Med Scand 221:179-184
Lawson DH, Jick H (1976) Drug prescribing in hospital: an international comparison.
Am J Public Health 66:644--648
Lehmann H, Ryan E (1956) Familial incidence of low pseudocholinesterase level.
Lancet 11:124
Levy M, Kewitz H, Altivein W, Hellebrand J, Eliakin S (1980) Hospital admissions due
to adverse drug reactions: a comparative study from Jerusalem and Berlin. Eur J
Clin PharmacoI17:25-31
Lindley CM, Tully MP, Paramsothy V, Tallis RC (1992) Inappropriate medication is a
major cause of adverse drug reactions in elderly patients. Age Ageing 21:294-300
Nolan L, O'Malley K (1988a) Prescribing for the elderly. I. Sensitivity of the elderly to
adverse drug reactions. J Am Geriatr Soc 36:142-149
Nolan L, O'Malley K (1988b) Prescribing for the elderly. II. Prescribing patterns:
differences due to age. J Am Geriatr Soc 36:245-254
Nolan L, O'Malley K (1989) The need for a more rational approach to drug prescribing
for elderly people in nursing homes. Age Ageing 18:52-56
Office of Health Economics (1992) Compendium of health statistics, 8th edn. Office of
Health Economics, London
Ouslander JG (1981) Drug therapy in the elderly. Ann Intern Med 95:711-722
Posner J, Rolan PE (1994) Clinical pharmacokinetics. In: Griffin JP, O'Grady J, Wells
FO (eds) The textbook of pharmaceutical medicine, 2nd edn. Greystone Books,
The Queen's University of Belfast, pp 333-351
Ramsay L (1981) The use of methyldopa in the elderly. J R ColI Physicians Lond
15:239-244
Ritschel WA (1980) Disposition of drugs in geriatric patients. Pharm Int 1:226-230
Roberts MA, Caird FI (1976) Steady-state kinetics of digoxin in the elderly. Age
Ageing 5:214
Royal College of Physicians (1984) Report on medication for the elderly. J Roy ColI
Physicians Lond 18:7-17
Scarfe MA, Israel MK (1994) Possible drug-interaction between warfarin and combina-
tion of levamisole and florouracil. Ann Pharmacother 28:464--467
Schenker S, Bay M (1994) Drug disposition and hepatotoxicity in the elderly. J Clin
Gastroenterol 18:232-237
Schmucker WD, Kontak JR (1990) Adverse drug reactions causing hospital admission
in an elderly population: experience with a decision algorithm. J Am Board Fam
Pract 3:105-109
304 J.e. McELNAY and P.F. D' ARCY: Age and Genetic Factors

Schneider JK, Mion LC, Frengley JD (1992) Adverse drug reactions in an elderly
outpatient population. Am J Hosp Pharm 49:90-96
Schwartz JT, Reuling FH, Feinlieb M, Garrison RJ, Collie DJ (1972) Twin heritability
study of the effect of corticosteroids on intraocular pressure. J Med Genet 9:137-
143
Scott PJ, Reid JL (1982) The effect of age on the response of human isolated arteries
to noradrenaline. Br J Clin Pharmacol 13:237-256
Shaw PG (1982) Common pitfalls in geriatric drug prescribing. Drugs 23:324-328
Shephard AM, Hewick DS, Moreland TA, Stevenson IH (1977) Age as a determinant
of sensitivity to warfarin. Br J Clin Pharmacol 4:315-320
Simons LA, Tett S, Simons J, Lauchlan R, McCallum J, Friedlander Y, Powell 1(1992)
Multiple medication use in the elderly - use of prescription and nonprescription
drugs in an Australian community setting. Med J Aust 157:242
Skaunic V, Hulek P, Martinkova K (1978) In: Kitani K (ed) Liver and ageing. Elsevier/
North Holland, Amsterdam, pp 115-130
Sloan RW (1992) Principles of drug therapy in geriatric patients. Am Fam Physician
45:2709-2718
Smith RL, Idle JR (1981) Genetic polymorphism in drug oxidation. In: Davis M,
Tredger JM, Williams E (eds) Drug reactions and the liver. Pitman, Bath, pp 95-
104
Stanton LA, Peterson GM, Rumble RH, Cooper GM, Polack AE (1994) Drug-related
admissions to an Australian hospital. J Clin Pharmacol Ther 19:341-347
Stewart RB, Cooper JW (1994) Polypharmacy in the aged. Practical solutions. Drugs
Aging 4:449-461
Sutherland FS, Carter JR (1975) Malignant hyperpyrexia during enflurane anaesthesia.
J Tenn Med Assoc 68:785-786
Telfer ABM, Macdonald DJZ, Dinwoodie AJ (1964) Familial sensitivity to
suxamethonium due to atypical pseudocholinesterase. Br Med J 1:153-156
Tinetti ME, Speechley M, Ginter SF (1988) Risk factors for falls among elderly persons
living in the community. N Engl J Med 319:1701-1707
Vesta RE, Wood AJ, Shand DG (1979) Reduced beta-adrenoreceptor sensitivity in the
elderly. Clin Pharmacol Ther 26:181-186
Vogel F (1959) Modem problems of human genetics. Ergeb Inn Med Kinderheilkd
12:52-125
Wallace DE, Watanabe AS (1977) Drug effects in geriatric patients. Drug Intell Clin
Pharm 11:597-603
Williams L, Lowenthal DT (1992) Drug therapy in the elderly. South Med J 85:127-131
Williamson J, Chopin JM (1980) Adverse reactions to prescribed drugs in the elderly:
a multicentre investigation. Age Aging 9:73-80
Woodhouse KW, Mutch E, Williams FM, Rawlins MD, James OF (1984) The effect of
age on pathways in drug metabolism in the human liver. Age Ageing 13:328-334
World Health Organization (1981) Health care in the elderly: report of the technical
group on the use of medicaments by the elderly. Drugs 22:279
CHAPTER 11
Drugs Causing Interference
with Laboratory Tests
S. YOSSELSON-SUPERSTINE

A. Introduction
Laboratory tests, along with the patient's history and the physical examina-
tion, often provide the main key to accurate diagnosis. In some cases their
abnormality is the only clue to diagnosis. Results of laboratory tests are also a
guide to rational therapy. They can reflect the effectiveness of the therapeutic
agent employed as well as indicate the appearance of adverse reactions. No
proper diagnosis and therapy can be provided without accurate and reliable
laboratory tests. Thus the sensitivity and the specificity of the test method
should be among the most important factors, more than cost and ease of
performance, in choosing and adopting a method or equipment. Specificity is
defined as the proportion of the true negatives that are correctly identified by
the test (ALTMAN and BLAND 1994) and is affected inversely by the number of
false-positive measurements as shown in the following formula (GADDIS and
GADDIS 1990):

CIty (0//0 ) =
SpecI·fi· True negatives x 100
True negatives + false positives
Many factors can cause a laboratory error or false-positive results. They
include human error, equipment or environmental changes, chemicals added
to specimens and the presence of endogenous substances. Another important
factor which has an effect is the presence of medications in the fluid tested
or influencing the functions in the body; this is then reflected in the test result.
This last factor - medications - is often overlooked. The mechanism of
their interference is the subject of this review. It is not the purpose of
this chapter to survey the literature and summarize the drug interferences
which have been documented with the various laboratory tests, nor is it
the aim to establish the clinical significance of drug-test interference. Only
examples of important interferences representing different mechanisms
will be discussed in more detail. The design of a study for the determination of
a clinically significant drug-test interaction resulting from methodological
interference will also be described. The reader is referred to excellent
compilations of published studies on this topic (SALWAY 1990; YOUNG 1990)
and to critical evaluations of some of the relevant literature (YOSSELSON-
SUPERSTINE 1986, 1989).
306 S. YOSSELSON-SUPERSTINE

B. Mechanisms of Drug-Test Interactions

In general, drug interferences can be classified into two major groups: (1)
pharmacological interferences, also known as biological or in vivo interfer-
ences and (2) methodological interferences, also known as analytical or in
vitro interferences.

I. Pharmacological Interferences
Pharmacological interferences are by far the most frequent type of interfer-
ence. They affect the result of the test by virtue of the activity of the drug or
its metabolites in the human body, regardless of the method employed in the
test. A pharmacological effect of a drug on a laboratory test is easy to detect
when the change in the test value is expected and wanted, but much more
difficult to interpret when it is an unexpected, toxicological affect of the drug.
The following examples demonstrate these effects:

1. Effect on Glncose Determination

When a hypoglycemic agent is administered to a diabetic patient, we expect
the results of the blood or urine glucose tests to be affected in a downward
direction. The sulfonylureas cause hypoglycemia by stimulating release of
insulin from pancreatic f3 cells and by increasing the sensitivity of peripheral
tissue to insulin (KAHN and SCHECHTER 1990). Therefore it is not surprising
that glipizide, a sulfonylurea agent, reduces glucose levels when given in a
therapeutic dose. However, when the pharmacological effect, detected in the
test, is not the reason for prescribing the drug, but rather an adverse reaction
to its administration, it is much more difficult to detect and interpret, espe-
cially when the reaction is not expected or necessarily seen in every patient or
in each course of therapy. Thus thiazide diuretics prescribed, for instance, to
treat hypertension, could elevate blood glucose from weeks to years after
therapy is started (JOSEPH and SCHUNA 1990) and the prevalence could be as
high as 30%. The adverse effect on glucose tolerance could be a result of
several possible mechanisms. One possibility is that the effect is not due to the
thiazide molecule itself, but to hypokalemia induced by the drug. Another
possibility is that thiazide compounds directly inhibit insulin secretion from
the f3 cells (BRASS 1984).

2. Effect on Uric Acid Determination

Another example is the pharmacological interference of drugs with uric acid
determination. When allopurinol is administered to a hyperuricemic patient in
a therapeutic dose it is expected to reduce uric acid levels. Both allopurinol
and its primary metabolite, alloxanthine, are inhibitors of xanthine oxidase -
an enzyme catalyzing oxidation of xanthines to form uric acid (INSEL 1990). On
the other hand, thiazides and loop diuretics cause uric acid elevation as a side
Drugs Causing Interference with Laboratory Tests 307

effect in as many as half the patients receiving them (ANON 1980). The preva-
lence of asymptomatic hyperuricemia could even rise to 65%-75% in elderly
patients being treated with diuretic, while the development of symptomatic
gout is commonly seen in only 1 %-2% of these individuals. Uric acid retention
begins soon after diuretic therapy is started, is dose dependent and produces
an average increase in serum uric acid concentration of 1.2-1.5 mgldl (WADE et
al. 1989). The precise mechanism is unclear. Several mechanisms have been
suggested, among them increased proximal tubular renal reabsorption, de-
creased tubular secretion and increased post-secretory reabsorption of uric
acid (MAY et al. 1992). The increase in uric acid will be demonstrated by all
laboratory methods for uric acid measurement, regardless of the technique
employed, further proof that the interference is pharmacological or toxicologi-
cal in its nature.

3. Intramuscular Injections and Muscle Enzyme Determination

Another type of drug-related interference in vivo is interference resulting
from the mode of drug administration, rather than the chemical moiety of the
drug. This is seen with the injection of medications intramuscularly, which
result in the elevation of muscle enzymes, such as creatine kinase (CK) or
aspartate aminotransferase (AST), and the possible erroneous attribution of
the rise to cardiac disease, mainly acute myocardial infarction, or to liver
disease. The rise in CK, for instance, could be of the order of two to six times
the normal concentration and lasting up to 48h postinjection (SEIFERT et al.
1992).

4. Effect of Drug-Drug Interactions

Drug-drug interactions are another mode of drug effect in vivo on a laboratory
test or an analysis employed to monitor the effect of another drug, or to
determine its blood concentrations. For example, there will be a dramatic
increase in prothrombin time values when phenylbutazone is added to war-
farin therapy as a result of combined effects of displacement of warfarin
from plasma protein-binding sites as well as inhibition of hepatic warfarin
metabolism. This is a pharmacokinetic interaction affecting the measurement
of prothrombin time (HANSTEN 1992). The administration of quinidine can
result in an increase in serum digoxin concentration of 0.5 ng/ml and more
(ANON 1994). The mechanism is variable and involves a quinidine-induced
30%-40% reduction in the apparent volume of distribution of digoxin and a
reduction of binding of digoxin in the cardiac muscle (SCHENCK-GUSTAFSSON
et al. 1981). Renal and nomenal clearances of digoxin are also reduced by
30%-40% (ANON 1994).
There is a very high variability in the outcome of this interaction as a result
of the numerous affecting factors, such as the sequence of administration,
duration of therapy, dose of drugs and other drugs taken by the patient. This
is why it remained undetected for half a century (HANSTEN 1992), and when
308 S. YOSSELSON-SUPERSTINE

first described was even thought to be an interference with the digoxin assay
methodology, which, if true, would not have necessitated an adjustment in the
digoxin dosage. However, this possibility was excluded by measuring digoxin
in plasma of patients on quinidine alone and finding no measurable concentra-
tions and by looking for an effect of quinidine on digoxin levels while adding
both drugs to plasma in vitro (EJVINSSON 1978).

II. Methodological Interferences

A drug or its metabolites can cause analytical error by interfering in the
various stages of the analysis, for instance, by competing with or imitating
the analyte participating in a chemical, immunological or enzymatic reaction;
by changing the chemical or physical conditions of the reaction environment;
or by falsely increasing the readings at the end of the analysis, such as color
or fluorescence measurements. It is important to study and document these
interferences and especially to research their clinical significance, so we can
choose a different method for a specific patient whose drug therapy cannot be
interrupted. There are alternatives to many of the laboratory tests available
today.

1. Colorimetric Interferences
Most of the laboratory tests in chemical pathology still have a colorimetric
component, which could be subject to the effect of foreign chromagens. The
following are examples of such interferences:

a) Uric Acid Analysis

The colorimetric method for the determination of uric acid in serum is an
example of a method in which drugs can react with the reagent and falsely add
to the reading of uric acid, if no modifications are made to eliminate reducing
substances. This method is still used in the SMAC Automated System
(TECHNIC ON SMAC 1976), and is based on the quantitative method of FOLIN
and DENIS (1912) as modified by MUSSER and ORTIGOZA (1966). In this method
phosphotungstate reagent is reduced by uric acid to phosphotungstite - a blue-
colored complex. Sodium tungstate is used as an alkalizing agent and hydroxy-
lamine is added to intensify the color. Drugs can interfere here if they also
reduce phosphotungstate. A survey of the literature (YOSSELSON-SUPERSTINE
et al. 1980; YOSSELSON-SUPERSTINE 1986) found that the most significant inter-
ferences were those with drugs such as acetaminophen, aspirin, ascorbic acid
and levodopa, which are commonly used and whose effect on uric acid deter-
mination has been studied in sera of volunteers or patients and not just in a
solution in vitro.
Acetaminophen increased serum uric acid by 17 % when a high dose of 2 g
was administered to 11 subjects (SINGH et al. 1972). In urine, uric acid in-
creased in concentration by 100%. No increase in serum uric acid concentra-
Drugs Causing Interference with Laboratory Tests 309

tion was noticed when the method of HENRY et al. (1957) was employed. This
method used carbonate-phosphotungstate and eliminated the use of cyanide
in the reaction used previously in order to reduce the turbidity of the final
colored solution, which made colorimetric measurement difficult. In addition
to modifying the test to eliminate acetaminophen interference, SINGH et al.
were also able to demonstrate a log-linear relationship between the concentra-
tion of the drug in aqueous solution and the concentration of the apparent uric
acid, further suggesting that the interference was with the method of uric acid
determination and not merely a pharmacological effect of the drug on uric
acid. CARAWAY (1969) demonstrated that ascorbic acid in a concentration of
IO.ug/ml, which could be found in the blood after the consumption of 2g of
the vitamin, increased serum uric acid concentration by 1.34mg/dl in
pooled sera and in sera of 21 patients, when determined by a carbonate-
phosphotungstate method. This interference is not significant in the more
sensitive cyanide methods but is of importance in methods where cyanide
is replaced by sodium silicate or carbonate (ALPER and SEITCHIK 1957).
Ascorbic acid was eliminated by mild alkaline treatment prior to adding
phosphotungstic acid (CARAWAY 1963). Incubation for 10min after adding
sodium carbonate effectively destroys the ascorbic acid present in serum and
urine in physiological concentrations.
High concentrations of uricase-resistant chromagens have been reported
in the serum of gouty patients receiving high maintenance doses of salicylate
(GRAYZEL et al. 1961). These metabolites have not been identified but were
not found to be gentisic acid - a salicylate metabolite interfering with uric acid
colorimetric assay in urine but not in plasma (CARAWAY 1969). The plasma
levels of salicylates in GRAYZEL'S study were above 13mg/lOOml. It should be
noted that at these levels there is also a pharmacological interference, a
uricosuric effect of salicylates, resulting in a lowering of uric acid levels. This
can affect the 7%-119% rise in uric acid as measured by a colorimetric method
(YO and GUTMAN 1959; GRAYZEL et al. 1961). On the other hand, the rise in
uric acid levels at salicylate doses of less than 2g/day is again caused by a
pharmacological effect of the drug leading to urate retention. Unconjugated
salicylate in the renal tubule may act by blocking a postulated tubular
secretion (YO and GUTMAN 1959). The distinction between the two types of
interferences - pharmacological and methodological - was possible by
measuring uric acid twice, once by the colorimetric assay and once by an
enzymatic spectrophotometric assay (GRAYZEL et al. 1961). These interfer-
ences are of great clinical significance since they can lead to an erroneous
diagnosis of gout in rheumatoid arthritic patients receiving various doses of
aspirin containing medications.
CAWEIN and HEWINS (1969) have demonstrated a 20% increase in serum
uric acid in 18 patients who received 3-7 g levodopa daily. This elevation was
absent when the uricase method was used. However, a pharmacological effect
cannot be ruled out since an elevation in serum uric acid, measured by the
uricase method, in two patients (AL-HuJAJ and SCHONTHAL 1971, 1972) and
310 S. YOSSELSON-SUPERSTINE

the appearance of symptoms of gout subsequent to the initiation of levodopa

treatment, were also reported (HONDA and GINDIN 1972). It should also be
noted that no methodological interference was found with therapeutic levels
of a-methyldopa - a chemically related substance, and uric acid determination
(SMALL et al. 1976). Combined enzymatic-colorimetric or spectrophotometric
methods for the determination of true uric acid, based on the destruction of
true uric acid by converting it to allantoin by uricase, and on the difference in
color development between total chromagens and nonurate chromagens,
should be employed whenever a drug interference with uric acid determina-
tion is suspected (CARAWAY and MARABLE 1966; MARTINEK 1970; SIGMA 1977;
KODAK EKTACHEM 1992a).

b) Creatinine Analysis
Creatinine measurement is one of the most valuable laboratory tools in clinical
practice. Not only does it provide an excellent estimation of the patient's
kidney function, but it also serves as a guide to dosage adjustments of drugs
whose elimination is mainly by the renal route. Two major techniques are used
today for the analysis of creatnine: colorimetric and enzymatic. The colorimet-
ric method is based on Jaffe's reaction, which was developed more than a
century ago and is still the most common assay for the determination of
creatinine either in serum or in urine (JAFFE 1886). It is employed in many
reagent kits and instruments, such as Beakman ASTRA-8, Technicon SMAC
and Du Pont ACA. In this reaction there is development of a red color
when creatinine reacts with picric acid in an alkali environment. The colored
end product absorbs light between 490 and 520nm and is measured by
spectrophotometer. The Jaffe reaction is efficient and inexpensive, but is
influenced not only by environmental conditions, and the amount of the
chemicals used, but also by noncreatinine chromagens, which could account
for 20% of the total measured creatinine in serum (NARAYANAN and ApPLETON
1980). These chromagens could be physiologic substances such as
glucose or protein (DATTA et al. 1986), as well as medications such as some of
the cephalosporin (GROTSCH and HAJDU 1987) and penicillin antibiotics
(KROLL et al. 1984a), lactulose (BRUNS 1988), high-dose furosemide infusion
(MURPHY et al. 1989), the today rarely used methyldopa (MADDOCKS et al.
1973; NANJI and WHITLOW 1984), acetohexamide (ROACH et al. 1985) and
phenacemide (RICHARDS 1980). Of the many cephalosporins studied, only
cefoxitin and cephalothin produced a significant interference at therapeutic
concentrations of 100mg/l (GREEN et al. 1990), as well as cafazolin, which
caused a false increase of 10-20,umol/l (0.1--D.2mg%) of creatinine for every
20mg/l cefazolin (NANJI et al. 1987).
The many publications on the interferences of these medications, and
especially cephalosporins, with the Jaffe analysis were evaluated very exten-
sively in a recent review (DUCHARME et al. 1993). The mechanism of interfer-
Drugs Causing Interference with Laboratory Tests 311

ence appears to be the presence of a carbonyl group on the offending com-

pound that alters the absorptivity of the complex (KROLL et al. 1987). Some of
the modifications suggested over the years, such as changing the pH or incuba-
tion temperature (LETELLIER and DESJARLAIS 1985a), usage of kaolin and
Lloyd's reagent (Fullers' earth) (HAECKEL 1981; BJERVE et al. 1988) and the
addition of a dialysis step could increase the specificity of the method. Thus,
for instance, a dialysis step diminished interference by cefoxitin using the
SMAC analytical system (KROLL et al. 1984b).
Timing of absorbance reading or usage of single-point kinetic methods
will further reduce drug interferences. The rate of color formation is different
for the cephalosporins and for creatinine. It reaches equilibrium in a shorter
period (8min) for creatinine (GROTSCH and HAJDU 1987).
The effect of timing of reading was also demonstrated in a study in which
creatinine was measured in the Beckman ASTRA 8, which performs two-
point kinetic reading, and the Technicon SMAC II, which makes a delayed
measurement. The ASTRA measurement exaggerated the apparent creati-
nine contribution from cephalothin (HICKMAN and MATHER 1988). However,
usage of the enzymatic assay (TOFFALETTI et al. 1983) is a better solution when
an unexplained high concentration of creatinine is measured in a patient with
an intact kidney, such as demonstrated many times by a normal blood urea
nitrogen (BUN) reading.
If creatinine is not determined by an enzymatic method in a patient
receiving a cephalosporin antibiotic, it is best that samples for creatinine
analysis be obtained at least 4-6h after infusion of the drug. The interference
is maximal when the sample is drawn within 30min after starting the infusion
(NANJI et al. 1987).
It is also recommended that creatinine be estimated by an enzymatic
method in patients on lactulose therapy since increases of3mg% and 6.Smg%
in creatinine, using the alkaline picrate assay, by the kinetic method (Beckman
ASTRA) and by the continuous flow method (Technicon SMAC) , respec-
tively, have been detected when a lactulose concentration of lOOg/1 was
present in the solution. The drug caused no problems with the enzymatic
method (BRUNS 1988). Likewise the enzymatic method is preferred when an
unusually low creatinine level is found, such as in cases of high-dose
furosemide consumption (MURPHY et al. 1989). In this case there is a possibility
of a metabolite blocking the colorimetric reaction since the parent drug did
not cause an interference in vitro. This error is of great clinical significance
because it could mask diminished renal function in patients receiving high
doses of furosemide (1-2g/dose).

c) Discoloration of Urine and Feces

Unusual color and appearance of urine or feces could provide useful
diagnostic information and should not be overlooked. For instance, red, pink
312 S. YOSSELSON-SUPERSTINE

or brown urine suggests hematuria, and orange or dark yellow urine suggests
the presence of urobilin. Bleeding from the upper gastrointestinal tract may
cause the stool to be black.
Drugs excreted in urine or feces can change the normal color and appear-
ance of these body wastes and obscure the elementary gross examination of
these specimens. Some also have the potential to affect tests performed on
urine or feces, especially when they are based on colorimetric methods.
Therefore, before proceeding with other diagnostic measures, a detailed
history of dietary intake of food and drugs should be obtained from the
patient. For instance, orally administered iron can cause false-positive reac-
tions for fecal occult blood when tested by the Hemoccult and Hematest
methods (LIFTON and KREISER 1982). Hemoccult uses guaic as a reagent and

Table 1. Drugs which may discolor urine

Drug/drug class Color produced

Acetanilide Yellow to red

Aloe Yellow-pink to red-brown in alkaline urine
Aminopyrine Red
Aminosalicylic acid Discoloration; red in hypochlorite solution
Amitriptyline Blue-green
Anisindione Pink or red to red-brown in acidic urine; orange in
alkaline urine
Antipyrine Red-brown
Azuresin Blue or green
Cascara Yellow-brown in acidic urine; yellow-pink in
alkaline urine, turning black on standing
Chloroquine Rust yellow to brown
Chlorzoxazone Orange or purplish red
Cimetidine (injection) Green
Clofazimine Red to brownish black
Danthron Pink to red or red-brown in alkaline urine
Daunonrubicin Red
Deferoxamine mesylate Reddish
Dimethylsulfoxide (DMSO) Reddish, due to hemoglobinuria
Diphenadione Orange in alkaline urine
Doxorubicin Red
Emodin Pink to red or red-brown in alkaline urine
Ethoxazena Orange to orange-brown
Ferrous salts Black
Furazolidone Rust yellow to brown
Idarubicin Red
Indigotindisulfonate Blue or green
Indomethacin Green due to biliverdinemia
Iron sorbitex Black
Levodopa Dark on standing in hypochlorite solution
Methocarbamol Dark to brown, black or green on standing
Methyldopa Dark on standing in hypochlorite solution
Methylene blue Blue or green
Metronidazole Dark
Mitoxantrone Dark blue or green
Drugs Causing Interference with Laboratory Tests 313

Table 1. Continued

Drug/drug class Color produced

Nitrofurantoin Rust yellow to brown

Pamaquine Rust yellow to brown
Phenacetin Dark brown to black on standing
Phenazopyridine Orange to orange-red
Phenindione Orange-red in alkaline urine
Phenolphthalein Pink to purplish red in alkaline urine
Phenolsulfonphthalein (PSP) Pink to red in alkaline urine
Phenothiazines Pink to red or red-brown
Phensuximide Pink to red or red-brown
Phenytoin Pink to red or red-brown
Primaquine Rust yellow to brown
Promethazine (injection) Green
Propofol (injection) Green
Quinacrine Deep yellow in acidic urine
Quinine Brown to black
Resorcinol Dark green
Riboflavin Yellow fluorescence
Rifampin Bright red-orange
Santonin Yellow in acidic urine, pink in alkaline urine
Senna Yellow-brown in acidic urine; pink to red in alkaline
urine; brown on standing
Sulfasalazine Orange-yellow in alkaline urine
Sulfonamides, antibacterial Rust yellow to brown
Thiazolsulfone Pink or red
Tolonium Blue-green
Triamterene Pale blue florescence
Warfarin Orange

Table 2. Drugs which may discolor feces

Drug/drug class Color produced

Antacids, aluminum hydroxide types Whitish or speckling

Antibiotics, oral Greenish gray
Anticoagulants, all Pink to red or black (if bleeding)
Bismuth-containing preparations Greenish black
Charcoal Black
Clofazimine Red to brownish black
Danthron Brownish staining of rectal mucosa
Dithiazanine Green to blue
Ferrous salts Black
Heparin Pink to red or black (if bleeding)
Indocyanine green Green
Indomethacin Green due to biliverdinemia
Nonsteroidal anti-inflammatory drugs Pink to red or black (if bleeding)
Phenazopyridine Orange-red
Pyrvinium pamoate Red
Rifampin Red-orange
Salicylates, especially aspirin Pink to red or black
Senna Yellow
314 S. YOSSELSON-SUPERSTINE

Hematest uses orthatolidone as a reagent. The reaction between these re-

agents and a substance acting as a peroxidase (hemoglobin) causes a blue-
colored reaction in the product tested. Other potential interactions with the
test causing false-positive results are with ascorbic acid, aspirin, povidone-
iodine and medications that irritate the gastroinetstinal tract, such as cortico-
steroids and nonsteroidal anti-inflammatory agents (ENGLE et al. 1988).
A comprehensive listing of drugs which may discolor feces and urine is
given in Tables 1 and 2 (ANON 1993).

2. Effect on pH of the Assay Environment

The pH of a solution is one of the major conditions for a specific chemical
reaction to be carried out. Many of the colorimetric reactions in laboratory
tests are pH dependent. Urine is one of the biological fluids whose pH is most
affected by diet and drugs. Acid urine may be produced by a diet high in meat
protein and in some fruits such as cranberries and by drug and chemicals, such
as ammonium chloride, methionine, methenamine mandelate, ascorbic acid or
acid phosphate. Alkaline urine may be induced by a diet high in certain fruits
and vegetables, especially citrus fruits and drugs such as sodium bicarbonate,
potassium citrate and acetazolamide (BRADLEY et al. 1979).
The following test is an example of a test affected by the change in urine
pH:

a) Urine Protein Reagent Strip Test

The colorimetric reagent strip test is based on the ability of proteins to alter
the color of some acid-base indicators without altering the pH. When an
indicator such as tetrabromophenol blue is buffered at pH 3 it is yellow in
solutions without protein, but in the presence of protein the color changes to
green, and then to blue with increasing protein concentrations.
The Albustix reagent strip is a protein test strip that contains a single
test area. This area consists of a small square of absorbant paper impregnated
with a buffered solution of tetrabromophenol blue. Uristix, N-Uristix,
Combistix, Hema-Combistix, Labstix, Bililabstix, Multistix and N-Multistix
and N-Multistix-SG reagent strips are multideterminant reagent strips, each
containing an area for protein determination along with test areas for other
urinary constituents. Protein is determined simply by dipping the strip into
well-mixed uncentrifuged urine, and immediately comparing the resultant
color with the chart provided on the reagent strip bottle. The results are
reported as negative (yellow color), trace, or one "plus" to four "plus". Trace
readings may detect protein in a concentration of 5-20mg/dl. "Plus" readings
are approximately equivalent to protein concentrations of 30, 100, 300 and
over 2000/dl, respectively, and are reliable indicators of increasingly severe
proteinuria. Albumin reacts with the indicator more strongly than do the other
proteins.
Today there is automated and semi-automated instrumentation using re-
Drugs Causing Interference with Laboratory Tests 315

fiector photometers that could read the reagent strips, allowing for elimination
of the variable that has commonly caused erroneous and confused results
when color has been interpreted by the human eye (ANON 1982).
Highly buffered, alkaline urines may give false-positive results when the
buffer systems in the reagent area are overcome and an actual shift in pH of
the buffers occurs. Addition of sodium carbonate (43 gil urine standard)
caused a +4 reading by the Albustix method (GYURE 1977).
Therapeutic doses of the commonly used antacids in over the counter
medications do not elevate urine pH by more than one unit (GIBALDI et al.
1974). Since normal pH varies from 4.6 to 8, it is unlikely that they affect urine
protein determination by the reagent strip method unless, at least theoreti-
cally, there is an abuse of antacid products.
Therapeutic doses of acetazolamide (37S-1000mg/day) have also not been
found to elevate urine pH to 10 or above and did not falsely elevate urine
protein levels (YOSSELSON-SUPERSTINE and SINAI 1986).

3. Interference with Chromatographic Methods

Drugs can interfere with various chromatographic methods. They, or their
metabolites, can confuse spots of analytes on thin layers or solvent chromato-
graphy or they can elute with the tested analyte using high-pressure liquid
chromatography, causing an erroneous increase in its concentration. The fol-
lowing are examples of such interferences:

a) Urine Amino Acid Screening

Urine screening for amino acid metabolic disorders can be performed effec-
tively by a semiquantitative chromatographic technique. First the amino acids
are separated by electrophoresis at acid pH, followed by solvent chromatogra-
phy for separation in the second dimension. Alternatively, thin-layer chroma-
tography can be used. Staining is performed by different reagents. Most alpha
amino acids and compounds containing primary or secondary amino groups
attached to an aliphatic carbon atom react with Ninhydrin (triketohydrindene
hydrate) reagent to form purple colors. Isatin reagent is particularly useful
for locating proline and hydroxyproline. Proline, which is yellow with
Ninhydrin reagent, reacts with isatin to form a blue color, which is much
more visible. Other amino acids also react to form unstable colors, varying
from blue to red-purple.
Urea and citrulline react with Ehrlich's reagent to form a yellow color.
This reagent is also useful for the location of tryptophan and indoles. Histi-
dine, carnosine and other imidazole derivatives react with Pauly's reagent to
form reddish colors. Hydroxyproline also reacts to form an orange-brown
color. Arginine and other monosubstituted guanidine derivatives react with
Sakaguchi's reagent to form an orange color.
Most sulfur-containing amino acids react with platinic-iodide reagent to
create white spots on a light pink background. With fast blue reagent,
316 S. YOSSELSON-SUPERSTINE

methylmalonic acid (MMA) appears as a deep-purple spot and ethylmalonic

acid (EMA) appears as a dark-blue/gray spot above MMA. Phenylpyruvic
acid and the branched-chain keto acids appear as bright-purple spots above
EMA. Homogentisic acid appears as a light-brown spot and lactic acid appears
as a bleached spot.
Several drugs or their metabolites can react with the mentioned reagents
and obscure the chromatographic results if they chromatograph near amino
acids or give false-positive results. A careful history of medications should,
therefore, be obtained when an unusual amino acid pattern is found in the
urine. In addition, authentic amino acids should be chromatographed with the
urine to show dissimilarity and a repeat urine amino acid screening should be
done after the patient has been off all medications for at least 3 or 4 days.
Drugs implicated in interfering with this test are ampicillin and other synthetic
penicillins, which appear as several purple and brown ill-defined spots to the
right of phenylalanine. Kanamycin antibiotic is excreted as a compound which
can react with Ninhydrin and which migrates faster than lysine. Carbenicillin
causes an ill-defined purple area at the right upper corner of the chromato-
gram, in addition to a purple spot to the right of phenylalanine and one to
the right of glutamine. Cephalexin gives a Ninhydrin-purple spot next to
phenylalanine. Leukemic patients receiving folate antimetabolite therapy
(e.g., methotrexate) and patients with vitamin B12 or folic acid deficiency may
excrete 4(5)-amino-5(4)-imidazole carboxamide (AIC), which has a mobility
similar to that of fi-aminoisobutyric acid, fi-alanine and o..aminolevulinic
acid and reacts with Ninhydrin to form a yellow-brown color. N-
Acetylcysteine, a mucolytic agent and an agent also used today to treat
acetaminophen poisoning, is excreted both in its oxidized form (N-
acetylcysteine) and as a mixed disulfide of N-acetylcysteine and cysteine. Both
derivatives react with the cyanide-nitroprusside reagent and with Ninhydrin
reagent (SHIH et al. 1991).

b) High-Pressure Liquid Chromatography Drug Assays

High-pressure liquid chromatography (HPLC) assays are possible on all thera-
peutically monitored drugs except lithium and digoxin. They can be developed
easily with no need for preparation of commercial reagent kits; however, they
are more costly and not without the problem of drug interference (YOSSELSON-
SUPERSTINE 1984, 1989; BOTTORFF and STEWART 1986). For instance, in one of
the HPLC methods which was developed to measure serum procainamide,
as well as its major active metabolite N-acetylprocainamide (NAPA),
sulfathiazole was found to interfere, when added in vitro, with the estimation
of NAPA (SHUKUR et al. 1977). In another HPLC method (DUTCHER and
STRONG 1977), a presumed metabolite of quinidine in therapeutic concentra-
tions (>3,ug/ml) caused the overestimation of NAPA by up to O.2,ug/ml.
Aminophylline, paracetamol (acetaminophen), codeine and caffeine were also
reported to be elevated by one of these methods (STEARNS 1981); however,
Drugs Causing Interference with Laboratory Tests 317

except for caffeine, which was injected in a concentration of 1.5,ug/ml (peak

plasma concentration after a dose of 100mg given as coffee), this happened
with toxic concentrations of these drugs. The interference of caffeine was
resolved in another HPLC procedure by changing the buffer concentration
(COYLE et al. 1987).
In another similar assay method (GANNON and PHILLIPS 1982),
metronidazole HCI in a dose of 500mg given intravenously every 6h coeluted
with procainamide and raised the baseline of the procainamide peak, thus
causing an erroneous increase in its blood concentration.
Of all the drugs commonly monitored in clinical practice, theophylline is
the one for which the largest list of reports has been published about assay
methodologies and clinically significant drug interferences. Most of these in-
terferences were reported with the use of HPLC methods, which are still used
(KELSEY et al. 1987) but have been replaced in many laboratories by immu-
noassays. Among the interfering drugs are many anti-infective agents such
as many of the cephalosporins, ampicillin, methicillin, sulfonamides,
metronidazole, chloramphenicol, analgesics such as salicylic acid and acetami-
nophen and many other drugs. The interference is dependent on the assay
conditions, such as type of mobile phase, pH adjustments and incorporation of
an ion-pair reagent. For a detailed list and discussion refer to YOSSELSON-
SUPERSTINE (1989). Because of their frequent coadministration it should be
emphasized that most of the interferences caused by cephalosporins are
clinically significant especially when the patient suffers from renal disease, in
which case an immunoassay method should be preferred (GANNON and LEVY
1984).

4. Interference with Enzymatic Reactions

Many of the laboratory tests used today include an enzymatic reaction step.
These reactions could be a part of a simple colorimetric method, as well as a
more sophisticated immunoassay. Drugs can interfere with these reactions by
activation of the reaction, by inhibition of the reaction, or by simply reacting
as the substrate and thus affecting the results of the analysis. The following
examples demonstrate these effects:

a) Creatinine Analysis
The enzymatic method for creatinine determination, which is used in the
automated drug-slide system, is based on the enzymatic hydrolysis of creati-
nine by creatinine iminohydrolase and the production of ammonia and N-
methylhydantoin. The ammonia reacts with bromophenol blue to form a blue
chromagen, the amount of which is measured spectrophotometrically. Its
concentration thus reflects the concentration of creatinine (TOFFALETII et al.
1983).
The 4-amino group of flucytosine, a systemic antifungal drug, can be
converted to ammonia by the same enzymatic system and cause false elevation
318 S. YOSSELSON-SUPERSTINE

of creatinine. Therapeutic plasma levels of flucytosine are between 400 and

775.umolll. When the flucytosine level is at the upper limit, the serum creati-
nine level may be falsely elevated by a factor of 14, whereas at sUbtherapeutic
flucytosine levels, such as 75.umolll, the serum creatinine level may be falsely
elevated by a factor of 2.5 (HERRINGTON et al. 1984; MITCHELL 1984; SOUNEY
and MARIANI 1985; O'NEILL et al. 1987). This interaction could be of much
clinical significance since flucytosine is frequently administered together
with amphotericin B and renal function is evaluated constantly along the
treatment period to detect any possible nephrotoxic reaction to the drug
therapy. The Ektachem 700 system of the second generation and above uses a
single-slide enzymatic method. It is more specific and does not cause false
elevation with flucytosine (COUTTIE et al. 1987). In this method the creatine
formed is converted to sarcosine and urea by creatine amidinohydrolase. The
sarcosine, in the presence of sarcosine oxidase, is oxidized to glycine, formal-
dehyde and hydrogen peroxide. The final step involves the peroxidase-
catalyzed oxidation of triarylimidazole leuko dye to produce a blue dye
(KODAK EKTACHEM 1992b).
This improved method did not completely eliminate the interference by
another drug, lidocaine (GOSNEY et al. 1987; SPINLER et al. 1989), and patients
on long-term lidocaine therapy may show an increase in creatinine concentra-
tion of up to 1.0mg/dl (88.umol/I). Most patients receiving intravenous
lidocaine will show an increase in creatinine concentration of O.3mg/dl
(26.umolll) or less (KODAK EKTACHEM 1992b). The interfering substance was
found to be N-ethylglycine (NEG), an inactive metabolite of lidocaine (SENA
et al. 1988). NEG is similar in structure to sarcosine (N-methylglycine) and can
thus serve as an additional substrate for sarcosine oxidase, producing false
elevation of creatinine. Because this metabolite is found in both serum and
urine the interference is noticed in creatinine determination in both urine and
serum. Until a further improvement in the single-slide method becomes avail-
able, serum creatinine should be analyzed by the picric acid method in patients
on lidocaine therapy.

b) Alkaline Phosphatase Analysis

In the determination of alkaline phosphatase, p-nitrophenyl phosphate, used
as the substrate, is hydrolyzed to p-nitrophenol at alkaline pH. p-Nitrophenol,
which is yellow, is read by a spectrophotometer. Alkaline phosphatase is the
catalyzing enzyme in this reaction and its concentration is correlated with the
concentration of the end product. Mg2+ is an enzyme activator in this reaction.
A drug metabolite such as cysteine (a metabolite of N-acetylcysteine) can
chelate with magnesium ion, causing a decrease in alkaline phosphatase
activity (LETELLIER and DESJARLAIS 1985b). Theophylline is another drug
that can affect this test result by directly inhibiting alkaline phosphatase
activity. It can cause a negative bias of up to 7% (KODAK EKTACHEM 1992c).
Two further drugs, methotrexate and nitrofurantoin, in five times therapeutic
Drugs Causing Interference with Laboratory Tests 319

concentrations, can falsely elevate alkaline phosphate levels. However, the

interference is spectral and not enzymatic in its nature (LETELLIER and
DESJARLAIS 1985b).

5. Interference with Immnnoassays

Immunoassays are gaining more and more acceptance as the method of choice
in the biochemical and in the clinical laboratory, for instance, in disposition
studies for biotechnology products or in therapeutic drug monitoring (TDM)
(BOTTORFF and STEWART 1986). Various types of immunoassays are employed,
the most frequently used being the homogeneous enzyme multiplied immuno-
technique (EMIT), enzyme-linked immunosorbent assay (ELISA), radioim-
munoassay (RIA) and fluorescence immunoassay (FIA). All of these assays
involve antigen-antibody reactions, so one important mechanism of drug in-
terference with these methods is that involving nonspecificity of antibodies.
This is demonstrated in the following assay:

a) Digoxin Assay
Digoxin is one of the most important drugs where the therapy is guided by its
concentrations in the patient's blood. All the immunoassays, radioimmunoas-
says, enzyme immunoassays and fluoroimmunoassays used for measurement
of the drug suffer from the cross-reactivity of digoxin metabolites and from
endogenous substances known as digoxin-like immunoreactive substances
(DLIs). These substances accumulate in the blood, especially in neonates,
during the third trimester of pregnancy, renal failure and liver disease (MORRIS
et al. 1990). The most important exogenous substance to interfere with digoxin
assays is the diuretic spironolactone - a synthetic steroid. Canrenone and its
20-hydroxy derivative, both metabolites of spironolactone, have apparently
more cross-reactivity with digoxin than the parent drug (HUFFMAN 1974;
SILBER et al. 1979; MORRIS et al. 1987). The wide range of falsely increased
levels of digoxin ranging from no effect to an increase of 4.0nglml after the
administration of therapeutic doses of spironolactone (25-200mg/day) can be
attributed to the difference in the specificity of the digoxin antiserum used in
the different RIA kits (MORRIS et al. 1987).
Endogenous steroids are known to cross-react with digoxin. An apparent
digoxin concentration of 0.5 mg/l has also been reported in the serum of a
patient receiving no digoxin, 30min after the administration of 6-methylpred-
nisolone (GAULT et al. 1985). The interference with steroids was noted with
both radioimmunoassay and with fluorescence polarization immunoassays
(SOLDIN et al. 1984). Fab fragments infused into patients with digoxin toxicity
pose another problem with digoxin radioimmunoassays, as well as with
fluorescence excitation transfer immunoassays (NATOWICZ and SHAW 1991).
Digoxin serum levels after Fab treatment could be as much as tenfold greater
than pretreatment values, when the solid-phase methods are used. The inter-
ference could be eliminated when polyethylene glycol and charcoal methods
320 S. YOSSELSON-SUPERSTINE

are used (GIBB et al. 1983). It is best not to use digoxin blood levels as a
guideline for continuation of the antidote therapy and to base the decision on
the clinical presentation of the patient, although the newer techniques for
measuring unbound digoxin could also be helpful (BANNER et al. 1992). It is
interesting to note that unconventional medicines could also participate in
drug-test interactions. This is important to remember when unusual results are
obtained since such medicines often do not appear on the patient's drug list.
Such an interference was noted with kyushin - a Chinese medicine that is very
popular in Japan, where it is obtainable without prescription. This medicine
was responsible for digoxin levels of more than 2.5 ng/ml in a patient who was
on a daily dose of 0.25 mg digoxin and did not exhibit any signs of digoxin
toxicity. The main components of the drug have chemical structures similar to
those of digoxin and digitalis-like cardiotonic actions. One tablet of kyushin
had a digoxin-like immunoreactivity equivalent to 1.9,ug (TDX analyzer,
Abbot Laboratories), 1.5,ug (Du Pont Aca V analyzer) and 72,ug digoxin
(Enzymun-Test, Boehringer Mannheim). The different equivalencies are
attributed to differences in cross-reactivity of the antibody used in the
immunoassays.
Two healthy volunteers took a typical dose of kyushin - two tablets three
times a day, and digoxin-like immunoreactivity reached almost OAng/ml in
half a day. The authors concluded that digoxin serum levels should be
interpreted very carefully in patients taking Chinese medicine (FUSHIMI et al.
1989).

c. Design of a Study for Evaluation

of Analytical Interference
The Expert Panel on Drug Effects in Clinical Chemistry (EPDECC) of the
International Federation of Clinical Chemistry (IFCC) suggested guidelines
for evaluation of analytical interference (GALTEAU and SIEST 1984). The first
step in the guidelines is to accumulate the relevant information (e.g., physico-
chemical and pharmacokinetic characteristics) about the drug, its metabolites
and formulations. Drugs to be selected for evaluation are those most
frequently used, those specific for the disease monitored by the laboratory test
studied or those suspected of interfering because of their physicochemical
characteristics.
The second step is to select the laboratory method to be studied and
choose a reference method of analysis. Once we have the drugs and the
methods, we are ready to start with in vitro studies on biological specimens,
serum, plasma or urine obtained from healthy subjects not on medication. We
prepare tenfold concentrated solutions of the drug and add to the serum,
plasma or urine pool. A control sample of equivalent volumes of the solvent
is also prepared and added to the biological specimens. The concentration
Drugs Causing Interference with Laboratory Tests 321

studied is the one which is ten times the highest therapeutic concentration
reported in the literature or, if unknown, the concentration calculated from
ten times the therapeutic daily dose diluted in 5 or 151, depending on the
volume of distribution of the drug. All the drug and control samples should be
tested in duplicate.
If an interference is established, we follow with studies to measure the
magnitude of the analytical interference. At least five different concentrations
are used with two of them in the therapeutic range. A dose-effect curve is
obtained and the slope calculated.
Further steps used in the various studies for the detection and evaluation
of a drug-laboratory method interaction are the repetition of the study using
many different assays after establishing the basic difference in the results
obtained by them, that is when no interfering substance is present. The studies
should then be carried out on volunteers' or patients' biological fluids after
the consumption of the potential interfering drug in various dosages or
concentrations.

D. Conclusions
In the present chapter, an attempt has been made to shed light on a very
important phenomenon, the effect of drugs on laboratory tests in clinical
practice. Laboratory tests are an indispensable tool for appropriate diagnosis
and therapy, and we need to be able to rely upon them. We must strive to
make them highly specific but, when this is not possible, the alternative is to
become familiar with the list of drug-test interferences. Knowledge of the
mechanisms of the interferences will assist us greatly in remembering and in
understanding them.
When we come across a suspected drug-test interference, there are certain
practical steps that we should follow. This is well presented in the suggested
algorithm by TROUB (1992). When the laboratory result is inconsistent with
other laboratory results or the clinical picture, the possibility of transcription
error should be ruled out and, if no such error is found, the test should be
repeated (with the new and original specimens). If the doubtful results are
consistent, a pharmacological or dietary effect should be ruled out. The litera-
ture should then be consulted for past documentation of possible analytical
interference. If this is found, either the offending drug should be discontinued
and the test repeated, or, if this is not possible, a different method of analysis
should be employed.
Whenever we come across a newly suspected drug-laboratory test inter-
ference, it should be made known to the medical community, so that others
will be able to perform further studies to evaluate the nature of the interfer-
ence and suggest ways to eliminate it. The documentation provided by such
studies will assist in minimizing possible errors in diagnosis, as well as in
treatment.
322 S. YOSSELSON-SUPERSTINE

References
Al Hujaj M, Schonthal H (1971) Hyperuricemia and levodopa. N Engl J Med 285:859-
860
AI-Hujaj M, Schonthal H (1972) Gout and levodopa. N Engl J Med 286:376
Alper C, Seitchik J (1957) Comparison of the Archibald-Ken and Stransky colorimetric
procedure and the Praetorius enzymatic procedure for the determination of uric
acid. Clin Chern 3:95-101
Altman DG, Bland JM (1974) Diagnostic tests 1: Sensitivity and specificity. Br Med J
308:1552
Anon (1980) Diuretics, hyperuricemia and tienilic acid. Lancet 11:681-682
Anon (1982) Proteins in urine. In: Modern urine chemistry, Ames Division, Miles
Laboratories, Elkhart, Indiana
Anon (1993) Drug-induced discoloration of feces and urine. In: Knoben E, Anderson
PO (eds) Handbook of clinical drug data. Drug Intelligence Publications,
Hamilton, Illinois, p 22
Anon (1994) Digoxin-quinidine. In: Tatro DS, Olin BR, Hebel SK (eds) Drug interac-
tion facts. Facts and comparisons, St. Louis, p 290
Banner JW, Bach P, Burke B, Freestone S, Gooch WM (1992) Influence of assay
methods on serum concentrations of digoxin during Fab fragment treatment. Clin
ToxicoI30:259-267
Bjerve KS, Egens J, Lampinen LM, Masson P (1988) Evaluation of several creatinine
methods in search of a suitable secondary reference method: report from the
subcommittee on reference method for creatinine. Nordic Society for Clinical
Chemistry. Scand J Clin Lab Invest 48:365-373
Bottorff MB, Stewart CF (1986) Analytical techniques and quality control. In: Taylor
WJ, Diers Caviness MH (eds) A textbook for the clinical application of therapeu-
tic drug monitoring. Abbott Laboratories, Texas, pp 51-57
Bradley M, Schuman GB, Ward PCJ (1979) Examination of urine. In: Henry JB (ed)
Clinical diagnosis and management by laboratory methods. WB Saunders Com-
pany, Philadelphia, pp 559-634
Brass EP (1984) Effects of hypertensive drugs on endocrine function. Drugs 27:447-
458
Bruns DE (1988) Lactulose interferes in the alkaline picrate assay for creatinine. Clin
Chern 34:2592-2593
Caraway WT (1963) Uric acid. Stand Meth Clin Chern 4:239-247
Caraway WT (1969) Non-urate chromagens in body fluids. Clin Chern 24:54-57
Caraway WT, Marable H (1966) Comparison of carbonate and uricase-carbonate
methods for the determination of uric acid in serum. Clin Chern 12:18-24
Cawein MJ, Hewins J (1969) False rise in serum uric acid after L-dopa. N Engl J Med
281:1489-1490
Couttie K, Earle J, Loakley J (1987) Evaluation of single-slide creatinine method on
the Kodak Ektachem 700 shows positive interference from lidocaine metabolites.
Clin Chern 33:1674
Coyle JD, Mackichan JJ, Boudoulas H, Lima JJ (1987) Reversed-phase liquid chroma-
tography method for measurement of procainamide and three metabolites in
serum and urine: percent of dose excreted as deethyl metabolites. J Pharm Sci
76:402-405
Datta P, Graham GA, Schoen I (1986) Interference by IgG paraproteins in the Jaffe
method for creatinine determination. Am J Clin Pathol 85:463-468
Ducharme MP, Smythe M, Strohs G (1993) Drug induced alterations in serum creati-
nine concentrations. Ann Pharmacother 27:622-633
Dutcher JS, Strong JM (1977) Determination of plasma procainamide and N-
acetylprocainamide concentration by high pressure liquid chromatography. Clin
Chern 23:1318-1320
Ejvinsson G (1978) Effect of quinidine on plasma concentrations of digoxin. Br Med J
1:279-280
Drugs Causing Interference with Laboratory Tests 323

Engle JP, Donnelly AJ, Lewis RK (1988) Fecal occult blood tests: potential drug
interactions. Consult Pharm 3:356-358
Folin 0, Denis W (1912) A new colorimetric method for the determination of uric acid
in blood. J BioI Chern 13:469-475
Fushimi R, Tachi J, Amino N, Miyai K (1989) Chinese medicine interfering with
digoxin immunoassays. Lancet 1:339
Gaddis GM, Gaddis ML (1990) Introduction to biostatistics: part 3. Sensitivity,
specificity, predictive value, and hypothesis testing. Ann Emerg Med 19:591-
597
Galteau MM, Siest G (1984) Drug effects in clinical chemistry. Part 2. Guidelines for
evaluation of analytical interference. J Clin Chern Clin Biochem 22:275-279
Gannon RH, Levy RM (1984) Interference of third-generation cephalosporins with
theophylline assay by high-performance liquid chromatography. Am J Hosp
Pharm 41:1185-1186
Gannon RH, Phillips LR (1982) Metronidazole interference with procainamide HPLC
assay. Am J Hosp Pharm 39:1966-1967
Gault MH, Longerich L, Dawe M, Vasder SC (1985) Combined liquid chromatogra-
phy/radioimmunoassay with improved specificity for serum digoxin. Clin Chern
31:1272-1277
Gibaldi M, Grundhofer B, Levy G (1974) Effect of antacids on pH of urine. Clin
Pharmacol Ther 16:520-525
Gibb I, Adams PC, Pamham AJ, Jenning K (1983) Plasma digoxin: assay anomalies in
Fab-treated patients. Br J Clin Pharmacol 16:445-447
Gosney K, Adachi-Kirkland J, Schiller HS (1987) Evaluation of lidocaine interference
in the Kodak Ektachem 700 Analyzer single-slide method for creatinine. Clin
Chern 33:2311
Grayzel AL, Liddle L, Seegmiller JE (1961) Diagnostic significance of hyperuricemia
in arthritis. N Engl J Med 265:763-768
Green AJE, Halloran SP, Mould GP (1990) Interference by newer cephalosporins in
current methods for measuring creatinine. Clin Chern 36:2139-2140
Grotsch H, Hajdu P (1987) Interference by the new antibiotic cefpirome and other
cephalosporins in clinical laboratory tests, with special regard to the "Jaffe" reac-
tions. J Clin Chern Clin Biochem 25:49-52
Gyure WL (1977) Comparison of several methods for semiquantitative determination
of urinary protein. Clin Chern 23:876-879
Hansten PD (1992) Drug interactions. In: Koda-Kimble MA, Young LY (eds) Applied
therapeutics, the clinical use of drugs. Applied Therapeutics, Vancouver, Wash-
ington, pp 2-1-2-12
Haeckel R (1981) Assay of creatinine in serum with use of Fuller's earth to remove
interferents. Clin Chern 27:179-183
Henry RJ, Sobel C, Kim J (1957) A modified carbonate-phosphotungstate method for
the determination of uric acid and comparison with the spectrophotometric
uricase method. Am J Clin PathoI28:152-160
Herrington D, Drusano GL, Smalls U, Standiford HC (1984) False elevation in serum
creatinine levels. JAMA 252:2962
Hickman PE, Mather P (1988) Variation among instruments in interference by cepha-
losporin in the Jaffe reaction for creatinine. Clin Chern 34:215-216
Honda H, Gindin RA (1972) Gout while receiving levodopa for parkinsonism. JAMA
219:55-57
Huffman DH (1974) The effect of spinonolactone and canrenone on digoxin radioim-
munoassay. Res Commun Chern Pathol Pharmacol 9:787
Insel P A (1990) Analgesic-antipyretics and antiinflammatory agents: drugs employed
in the treatment of rheumatoid arthritis and gout. In: Gilman AG, RaIl TW, Nies
AS, Taylor P (eds) Goodman and Gilman's The pharmacological basis of thera-
peutics. Pergamon Press, New York, pp 638-681
Jaffe M (1886) Uber den Niederschlag we1chen Pikrinsaeure in normalen Ham erzeugt
und tiber eine neue Reaktion des Kreatinins. Z Physiol Chern 10:391-400
324 S. YOSSELSON-SUPERSTINE

Joseph JC, Schuna AA (1990) Management of hypertension in the diabetic patient.

Clin Pharm 9:864-873
Kahn CR, Schechter Y (1990) Insulin, oral hypoglycemic agents, and the pharmacology
of the endocrine pancreas. In: Gilman AG, RaIl TW, Nies AS, Taylor P (eds)
Goodman and Gilman's The pharmacological basis of therapeutics. Pergamon
Press, New York, pp 1463-1495
Kelsey HC, Ball MJ, Kay JDS (1987) Interference by acetazolamide in theophylline
assay depends on the method. Lancet 11:403
Kodak Ektachem (1992a) Test methodology notebook. URIC. Eastman Kodak Com-
pany, Rochester, New York
Kodak Ektachem (1992b) Test methodology notebook. CREA. Eastman Kodak Com-
pany, Rochester, New York
Kodak Ektachem (1992c) Test methodology notebook. ALKP. Eastman Kodak Com-
pany, Rochester, New York
Kroll MH, Hagengruber C, Elin RJ (1984a) Reaction of picrate with creatinine and
cepha-antibiotics. Clin Chern 30:1664-1666
Kroll MH, Hagengruber C, Elin RJ (1984b) Effect of dialysis on interference by
cefoxitin with determination of creatinine. Clin Chern 30:1386-1388
Kroll MH, Roach NA, Poe B, Elin J (1987) Mechanism of interference with the Jaffe
reaction for creatinine. Clin Chern 33:1129-1132
Letellier G, Desjarlais F (1985a) Analytical interference of drugs in clinical chemistry
II - The interference of three cephalosporins with the determination of serum
creatinine concentration by the Jaffe reaction. Clin Biochem 18:352-356
Letellier G, Desjarlais F (1985b) Analytical interference of drugs in clinical chemistry:
1 study of twenty drugs on seven different instruments. Clin Biochem 18:345-
351
Lifton U, Kreiser J (1982) False positive stool occult blood tests caused by iron
preparations: a controlled study and a review of the literature. Gastroenterology
83:860-863
Maddocks J, Hann S, Hospkins M, Coles GA (1973) Effect of methyldopa on creati-
nine estimation. Lancet 1:157
Martinek RG (1970) Review of methods for determining uric acid in biologic fluids.
JAm Med TechnoI32:233-241
May DB, Young LY, Wiser TH (1992) Essential hypertenstion. In: Koda-Kimble MA,
Young LY (eds) Applied therapeutics, the clinical use of drugs. Applied Thera-
peutics, Vancouver, pp 7-1-7-32
Mitchell EK (1984) Flucytosine and false elevation of serum creatinine level. Ann
Intern Med 101:278
Morris RG, Lognado PY, Lehmann DR, Fremin DB, Glistak ML, Bunret RB (1987)
Spironolactone as a source of interference in commercial digoxin immunoassay.
Ther Drug Monit 9:208-211
Morris RG, Frewin DB, Saccoia NC, Goldsworthy WL, Jeffries WS, McPhee AJ (1990)
Interference from digoxin-like immunoreactive substance(s) in commercial
digoxin kit assay methods. Eur J Clin Pharmacol 39:359-363
Murphy J, Hurt TL, Griswold WR, Peterson BM, Rodarte A, Krous HF, Reznik VM,
Mendoza SA (1989) Interference with creatinine concentration measurement by
high dose furosemide infusion. Crit Care Med 17:889-890
Musser W A, Ortigoza C (1966) Automated determination of uric acid by the hydroxyl-
amine method. Tech Bull Reg Med Tech 30:21-25
Nanji AA, Whitlow KJ (1984) Spurious increase in serum creatinine associated with
intravenous methyldopa therapy. Drug Intell Clin Pharm 18:896-897
Nanji AA, Poon R, Hinherg I (1987) Interference by cephalosporins with creatinine
measurement by desk-top analysers. Eur J Clin PharmacoI33:427-429
Narayanan S, Appleton HD (1980) Creatinine: a review. Clin Chern 26:1119-1126
Natowicz M, Shaw L (1991) Digoxin assay anomalies due to digoxin specific Fab
immunotherapy. Drug Intell Clin Pharm 25:739-741
O'Neill MB, Hill JG, Arbus GS (1987) False elevation of the serum creatinine level
associated with the use of flucytosine. Can Med Assoc J 137:133-134
Drugs Causing Interference with Laboratory Tests 325

Richards RK (1980) Structure-activity relationships in the effects of phenacemide

analogs on serum creatinine and anticonvulsant activity. Arch Int Pharmacodyn
244:107-112
Roach NA, Kroll MH, Elin RJ (1985) Interference by sulfonylurea drugs with the Jaffe
method for creatinine. Clin Chim Acta 151:301-305
Salway JG (1990) Drug-test interactions handbook. Chapman and Hall Medical,
London
Schenck-Gustafsson K, Jogestrand T, Nordlander R, Dahlquist R (1981) Effect of
quinidine on digoxin concentration in skeletal muscle and serum in patients with
atrial fibrillation. N Engl J Med 305:209-211
Seifert CF, Bradberry JC, Resman-Targoff BH (1992) Clinical laboratory tests and
interpretation. In: Herfindal ET, Gourley DR, Hart LL (eds) Clinical pharmacy
and therapeutics. Williams & Wilkins, Baltimore, pp 61-81
Sena SF, Syed D, Romeo R, Krzymowski GA, McComb RB (1980) Lidocaine metabo-
lite and creatinine measurements in the Ektachem 700: steps to minimize its
impact on patient care. Clin Chern 34:2144-2148
Shih VE, Mandell R, Sheinhait 1(1991) General metabolic screening tests. In: Hommes
FA (ed) Techniques in diagnostic human biochemical genetics. Willey-Liss, New
York, pp 61-81
Shukur LR, Powers JL, Marques RA, Winter ME, Sadee W (1977) Measurement of
procainamide and N-acetylprocainamide in serum by high-performance liquid
chromatography. Clin Chern 23:636-638
Sigma Chemical Co (1977) The enzymatic-calorimetric determination of uric acid in
serum or urine. Technical Bulletin, No. 680, St. Louis
Silber B, Sheiner LB, Powers JL, Winter ME, Sadee W (1979) Spironolactone-associ-
ated digoxin assay interference. Clin Chern 25:48-50
Singh HP, Hebert MA, Gault MH (1972) Effect of some drugs on clinical laboratory
values as determined by the Technicon SMA 12/60. Clin Chern 18:137-144
Small RE, Fredy HR, Small BJ (1976) Alpha-methyldopa interference with the
phosphotungstate uric acid test. Am J Hosp Pharm 33:556-560
Soldin SJ, Papanastasiou-Diamandi A, Heyes J, Lingwood C, Olley P (1984) Are
immunoassays for digoxin reliable? Clin Biochem 17:317-320
Souney PF, Mariani G (1985) Effect of various concentrations of fiucytosine on the
accuracy of serum creatinine determinations. Am J Hosp Pharm 42:621-622
SpinIer SA, Anderson BD, Kindwall KE (1989) Lidocaine interference with Ektachem
Analyzer determinations of serum creatinine concentration. Clin Pharm 8:659-
663
Stearns FM (1981) Determination of procainamide and N-acetylprocainamide by high-
performance liquid chromatography. Clin Chern 27:2064-2067
Technicon Instrument Corporation (1976) Technicon SMAC high-speed computer-
controlled biochemical analyzer, product labelling. Technicon Instrument Corpo-
ration, New York
Toffaletti J, Blosser N, Hall T, Smith S, Tompkins D (1983) An automated drug-slide
enzymatic method evaluated for measurement of creatinine in serum. Clin Chern
29:684-687
Troub SL (1992) Evaluating potential drug interferences with laboratory tests. In:
Troub SL (ed) Basic skills in interpreting laboratory data. American Society of
Hospital Pharmacists, Bethesda, pp 11-21
Wade WE, Waite WW, Cobb HH (1989) Hyperuricemia associated with thiazide
diuretic therapy in an ambulatory geriatric hypertensive population. Consult
Pharm 4:94-98
Yosselson-Superstine S (1984) Drug interferences with plasma assays in therapeutic
drug monitoring. Clin Pharmacokinet 9:67-87
Yosselson-Superstine S (1986) Drugs causing interference with laboratory tests. In:
D'Arcy PF, Griffin JP (eds) Iatrogenic diseases, 3rd edn. Oxford University Press,
Oxford, pp 954-969
Yosselson-Superstine S (1989) Drug interferences in therapeutic drug monitoring.
Giorn It Chim Clin 14:23-34
326 S. YOSSELSON-SUPERSTINE: Drugs Causing Interference with Laboratory Tests

Yosselson-Superstine S, Sinai Y (1986) Drug interference with urine protein determi-

nation. J Clin Chern Clin Biochem 24:103-106
Yosselson-Superstine S, Granit D, Superstine E (1980) Drug interference with the
phosphotungstate uric acid test. Am J Hosp Pharm 37:1458-1462
Young DS (1990) Effects of drugs on clinical laboratory tests. AACC, Washington
Yii TF, Gutman AB (1959) Study of the paradoxical effects of salicylate in low,
intermediate and high dosage on the renal mechanisms of excretion of urate in
man. J Clin Invest 38:1298-1315
CHAPTER 12
Drug Interactions with Herbal
and Other Non-orthodox Remedies
P.A.G.M. DE SMET and P.F. D'ARCY

A. Introduction
Medicines derived from plants formed the majority of the earlier materia
medica because chemically synthesised compounds were then not available.
Many of these herbs have stood the test of time and critical clinical assessment
and have found their way into the pharmacopoeias of orthodox medicines
sometimes as the isolated and chemically standardised active ingredient. Such
drugs as cocaine, colchicine, coumarin anticoagulants, digoxin, ephedrine,
morphine, quinine and quinidine, reserpine, tubocurarine, sennosides, and the
ergot and vinca alkaloids entered orthodox medicinal use by this route.
There are many types of herbal remedies, ranging from self-made teas
prepared from self-collected herbs to officially registered drug products which
have passed through the same rigid registration procedure as synthetic medi-
cines (see Table 1).
It is rather difficult to classify a particular herb in its most appropriate
category. The same botanical product which is a registered medicine in one
country may be a dietary supplement or recreational herb in the next, and in
one country the same herb may be available as an official medicine, as a
health food preparation and as a raw ingredient (DE SMET 1993a). Yet it is
important to keep these different categories in mind, when the risks of herbal
remedies are discussed, because the nature and magnitude of these risks can
vary considerably with the specific type of product. For instance, a consumer
of a high-quality registered herbal medicine should not have to worry about
the correct identity of the ingredients, whereas this should be a primary
concern for an individual who goes out into the field to collect his own herbal
materials.
The majority of preparations used in herbal or non-orthodox medicines
are a mixture of herbal ingredients. Herbs having diverse actions may be
incorporated into one concoction. Also herbal products are, in some circum-
stances, a mixture of herbal remedies with other ingredients of non-herbal
origin (e.g., arsenic or lead) or undeclared Western drugs (e.g., prednisolone,
non-steroidal anti-inflammatory/antirheumatic agents and paracetamol) as
found recently by KARUNANITHY and SUMITA (1991) in traditional Chinese
antirheumatic medicines. Readers are also referred to the reviews of D'ARCY
(1991,1993) and to a comprehensive review by DE SMET (1992a) on adultera-
328 P.A.G.M. DE SMET and P.F. D'ARCY

Table 1. Different types of herbal remedies

Raw materials for self-preparation

- By self-collection
- Through commercial channels
(Semi)finished non-medicinal products
- Dietary supplements
- Health foods
- Recreational herbs, etc,
Registered medicines
- By special procedure
- By regular procedure

tion and contamination of herbal products with toxic metals and synthetic drug
substances and other drugs used in non-orthodox medicine.
It must be appreciated that the quality control exerted over most herbal
preparations and other non-orthodox remedies which are not registered as
medicines is often poor and more likely to be non-existent and that most
preparations are not standardised for potency in biological test systems. As a
consequence their potency may vary considerably from sample to sample.
Some of the interactions indicated in this review are based on a firm
pharmacological basis but their clinical relevance still needs to be established.
Others have been presented in anecdotal reports with lack of data essential to
establish a firm cause-effect relationship. In conformity with the objective of
this volume, mechanisms of the interaction are given where this is possible;
that it is not always possible is acknowledged since often the available data do
not permit even an assumption of the cause.
This approach in the context of this chapter is necessary if warnings are to
be given about likely interactions since the sparse reports of drug interactions
between non-orthodox products and Western medicines neither confirms their
safety in use nor indeed suggests that the incidence of such interactions is low.
The simple fact is that most interactions of this type will not be recognised as
such by the self-medicated patient and will not be reported to an orthodox
medical practitioner and therefore will not appear in the medical or pharma-
ceuticalliterature.
When such original reports of reactions and interactions, as are available,
are carefully analysed, it becomes obvious that many of the cases where herbal
products have been associated with actual human poisoning were not in fact
caused by the herbs alleged to be in the product, but resulted from substitution
or contamination of the declared ingredients, intentionally or by accident, with
a more toxic botanical, a poisonous metal, or a potent non-herbal drug sub-
stance (DE SMET 1992a).
Although herbal medicines are by far the largest component of non-
orthodox remedies they do not have exclusive claims; it must be made clear
that there are various types of other alternative treatments ranging from
preparations of animal origin, minerals, vitamins and amino acids. Many of
Drug Interactions with Herbal and Other Non-orthodox Remedies 329

these are capable of interfering with orthodox medicines. Interactions be-

tween minerals and vitamins and orthodox medicines have already been excel-
lently reviewed in standard drug interaction texts (GRIFFIN et al. 1988;
STOCKLEY 1991; HANSTEN et al. 1993) and therefore this present chapter will
largely confine itself to interactions involving herbals, animal origin prepara-
tions and amino acids. It is accepted that firm evidence for many of the
interactions is slight but it is thought better at this stage of knowledge to
include them and give warnings rather than neglect them to the possible
detriment of the medicated patient.
For convenience, and to illustrate the points that various types of alterna-
tive treatments can interfere with orthodox drugs, the subject will be discussed
under the following headings: animal agents; amino acids; vitamins and miner-
als; dietary fads; and, in much more detail, herbal medicines (DE SMET 1992b).

B. Animal Agents
There are few parts of animals that have not been used at one time or another
for their medicinal properties. For example, crude thyroid hormones may
occur as ingredients of slimming preparations and/or adulterants of herbal
products (DE SMET 1992a).

I. Fish Oil
Fish oil can upset coagulation control by increasing the bleeding time and
reducing platelet aggregation. It is difficult therefore to exclude the possibility
of untoward consequences in stabilised anticoagulated patients if this oil is
taken concomitantly with anticoagulant medication (DE SMET 1989).

II. Chinese Toad Venom

A traditional component of the Chinese medicine "Kyushin" is the venom of
the Chinese toad (Bufo bufo gargarizans). The venom contains bufalin and
cinobufaginal, which are chemically similar to digoxin. The venom may there-
fore interfere with digoxin immunoassays; the constituents can react with
digoxin antibodies and create the false impression of high plasma digoxin
levels (FUSHIMI et al. 1989, 1990; LIN et al. 1989; KWAN et al. 1992; PANESAR
1992).

c. Amino Acids

I. L- Tryptophan

L-Tryptophan (LT) is used in the orthodox treatment of depressed patients not

responsive to other therapy and also as a sedative-hypnotic. In non-orthodox
330 P.A.G.M. DE SMET and P.F. D'ARCY

treatments, LT is a common health food supplement. It is capable of interact-

ing with other serotonergic drugs, and combinations of LT with monoamine
oxidase inhibitor (MAOI) or fluoxetine have repeatedly induced a so-called
serotonin syndrome which is characterised by neurobehavioural problems
such as myoclonus, shivering, diaphoresis, hyperthermia, hyperreflexia, ocular
oscillations, ataxia and/or hypomanic changes (DE SMET 1991).
A marked and rapid deterioration has been observed in parkinsonian
patients when 100mg pyridoxine (vitamin B6)/day was added to a daily regi-
men of 8-9g LT. This deleterious effect was not produced by pyrodoxine or
LT alone (DE SMET 1991). Pyridoxine also reduces the effect of levodopa
in the treatment of parkinsonism (DUVOISIN et al. 1969). Pyridoxine is a
code carboxylase and by facilitating the decorboxylation of levodopa it reduces
its blood level. Another reported interaction is the significant reduction of
blood concentrations of levodopa when LT is given concomitantly (DE SMET
1991).

D. Vitamins
Vitamin and mineral supplements are a common combination and are readily
obtainable as over-the-counter medicines; they may be regarded by consumers
as food supplements rather than medication and the possibility of their inter-
action with prescribed medicines may not be appreciated. Many drugs are
known to evoke vitamin deficiency states, for example isoniazid, hydralazine
and penicillamine are antagonistic to vitamin B6; anticonvulsants cause folate
and vitamin D deficiency and have been implicated in osteomalacia (DENT et
al. 1970); and vitamin B12 deficiency may occur after prolonged treatment with
the antidiabetic metformin (CALLAGHAN et al. 1980; ADAMS et al. 1983).
They can influence the effects of various orthodox drugs and sometimes
this may be used intentionally as an advantage (see SOKOL et al. 1991), but
more often unwanted effects occur. For example health foods and food
supplements containing appreciable quantities of vitamin K can reduce the
effect of oral anticoagulants (HEALD and POLLER 1974; STOCKLEY 1991). Cases
of impaired anticoagulation have been recorded after concomitant intake of
"Gon" (a non-orthodox remedy containing vitamin K4), enteral nutrition, and
excessive amounts of green vegetables (UDALL and KROCK 1968; HEALD and
POLLER 1974; HOGAN 1983; KEMPIN 1983; WALKER 1984; WATSON et al. 1984).
Vitamin B6 can nullify the beneficial effects of levodopa in parkinsonism
(COTZIAS 1969; CALNE and SANDLER 1970). A further example is that folic acid
may occasionally decrease serum phenytoin to a clinically significant degree
(HANSTEN et al. 1993). Vitamins may also influence the activity of other non-
orthodox remedies, for example the absorption of selenium from sodium
selenite can be drastically reduced by megadoses of vitamin C (ROBINSON et al.
1985), and the use of large doses of vitamin C (up to 5 g/day), taken prophylac-
tically against the common cold, may alter urinary pH sufficiently to influence
Drug Interactions with Herbal and Other Non-orthodox Remedies 331

the renal elimination and therefore the clinical response of alkaline drugs
(MEDICAL LETTER 1971).
The amount of vitamin consumed is dependent on the specific product and
its specific dose recommendations. While many multivitamin preparations
provide daily amounts below or just above their Recommended Daily Allow-
ance (RDA) values, there are also products which provide megadoses of one
or more vitamins, This latter category entails a risk of toxic effects (EVANS and
LACEY 1986; BROWN and GREENWOOD 1987; HELSING 1992) as well as drug
interactions. In general, however, with the majority of multivitamin/mineral
supplements, it is doubtful whether the amount of vitamin contained in the
product plays any great part in interactions with orthodox drugs.

E. Minerals
It is normally the mineral salt component of the combination as well as the
mineral content of antacids and some laxatives that is responsible for the
majority of interactions. For example, the absorption of tetracyclines is re-
duced by the formation of insoluble chelates in the gut in the presence of milky
and other foods containing bivalent and trivalent metal ions (Ca, Mg, AI, Fe)
(KUNIN and FINLAND 1961; BRAYBROOKS et a1. 1975; CHIN and LACH 1975), and
serum levels of oral tetracyclines may be decreased by more than 50% if they
are taken with ferrous sulphate or milk (NEUVONEN 1976).
Iron supplements are among the most frequently prescribed drugs (LA
PlANA SIMONSEN 1989) and are taken in over-the-counter remedies by many
other people. CAMPBELL and HASINOFF (1991) have listed some 19 orally
administered drugs, including antimicrobials, captopril, levodopa, salicylic
acid and thyroxine, which have functional groups in a configuration that will
bind iron to form an iron-drug complex. The formation of a stable iron-drug
complex reduces the extent of drug absorption but does not appear to reduce
the rate of drug absorption. Iron can also catalyse oxidation and reduction
interactions. Zinc also can produce various interactions with orthodox drugs
(BARRY et a1. 1991).
Materials with a high calcium content may reduce the absorption of the
bisphosphonates. Vitamins with mineral supplements such as iron, calcium
supplements, laxatives containing magnesium, or antacids containing calcium
or aluminium all dramatically reduce the biological availability of the
bisphosphonates (FLEISCH 1987; FELS et a1. 1989).

F. Dietary Fads
Dietary fads may lead to an excessive intake of certain nutrients, which in turn
could result in interference with orthodox medication. Certain diets, including
vegetable-rich weight-reducing diets, are rich in vitamin K and if taken in
sufficient amounts will interfere with warfarin treatment (HEALD and POLLER
332 P.A.G.M. DE SMET and P.F. D'ARCY

1974; HOGAN 1983; PATRIACA et al. 1983; WALKER 1984; WATSON et al. 1984).
For example, a patient maintained on warfarin 8mg daily required an increase
in dosage to 13 mg/day when placed on enteral nutrition rich in vitamin K; the
warfarin dosage returned to 8mg/day when the food supplement was stopped.
Problems have also been experienced with anticoagulated patients consuming
large amounts of broccoli (KEMPIN 1983; STOCKLEY 1984).
Beside the risk of pharmacodynamic problems, there is also the possibility
that the pharmacokinetic fate of orthodox drugs could be affected by a dietary
change. For instance, dietary fibres may affect absorption (see below),
whereas certain other foodstuffs may exert an influence on drug biotransfor-
mations in humans (ALVARES 1984; ANDERSON 1988). Also certain foods and
food supplements can affect the urinary excretion of drugs. A balanced diet
with adequate protein provides an acid urine, whereas a low-protein diet or a
strict vegetarian diet may give an alkaline urine. This opens up the possibility
that diet-induced changes in urinary pH affect the rates of excretion of weakly
acidic or weakly basic drugs (WESLEY-HADZIJA 1971; PIERPAOLI 1972). Acid
urine favours the ionisation of alkaline drugs (and vice versa), so reabsorption
is reduced and renal excretion is increased. Acid fruit juices and squashes may
therefore reduce the efficacy of the antimalarials quinine and chloroquine
(P.F. D'ARCY, personal observation).
Grapefruit juice, but not orange juice, greatly augments the bioavailability
of the antihypertensive, calcium-antagonists felodipine, nifedipine and
nitrendipine (BAILY et al. 1989, 1991; SOONS et al. 1991; EDGAR et al. 1992);
similar results are apparent on the clearance of caffeine (FUHR et al. 1993).
These affects are due to the inhibition of cytochrome P450 (isoforms CYP1A2
and CYP3A4) enzymes by several flavonoid glycosides present in grapefruit
juice, of which the bitter principle naringin (4',5,7-trihydroxyflavanone 7-
rhamno-glucoside) is the most abundant (FUHR et al. 1993).
Interestingly, grapefruit juice has also recently been found to inhibit the 7-
hydroxylation of coumarin, as measured by its effect on urinary excretion of 7-
hydroxycoumarin in healthy volunteers (MERKEL et al. 1994). This implies that
grapefruit juice may interfere with the pharmacokinetics of coumarin-yielding
medicinal herbs, such as Melilotus officinalis (sweet clover), Asperula odorata
(sweet woodruff), Dipteryx odorata (tonka bean) and Anthoxanthum
odoratum (sweet vernal grass) (HOPPE 1975; LEWIS and ELVIN-LEWIS 1977;
TEUSCHER and LINDEQUIST 1980). A comprehensive review on the significance
of interactions between grapefruit juice and drugs was recently presented by
BAILEY et al. (1994).

G. Herbal Drugs
This category includes crude herbal remedies as well as plant-derived drugs;
discussion will be divided into the following categories:
Drug Interactions with Herbal and Other Non-orthodox Remedies 333

1. Effects of herbal drugs on orthodox drug pharmacokinetics: effects on

absorption
2. Effects of herbal drugs on orthodox drug pharmacokinetics: effects on
elimination
3. Effects of orthodox drugs on herbal drug pharmacokinetics
4. Pharmacodynamic interactions between herbal drugs and orthodox drugs
5. Multiple or unclarified interactions between herbal drugs and orthodox
drugs
6. Interactions between different herbal drugs

I. Effects of Herbal Drugs on Orthodox Drug Pharmacokinetics:

Effects on Absorption

1. I>ietary Fibres
BROWN et al. (1977) showed that digoxin bioavailability is decreased by almost
20% when given with a high-fibre meal. In vitro studies by FLOYD et al. (1977)
supported this and indicated that up to 45 % of the digoxin may be sequestered
in or bound to the bran. Later work by REISSELL and MANNINEN (1982)
showed that the effect of fibre was small but that an interactive effect on
absorption of digoxin might occur if fibre and digoxin were ingested simulta-
neously.
ROSSANDER (1987) suggested, on the basis of in vitro studies, that dietary
fibres might interfere with the bioavailability of iron in the diet. Confirmatory
work in healthy volunteers showed that bran had a marked inhibitory effect on
the absorption of dietary iron but not cellulose or pectins. ROE et al. (1988)
showed that dietary fibre in combined wheat bran and psyllium biscuits or
biscuits containing psyllium alone reduced the absorption of riboflavin by
6.4% and 5.7%, respectively. No effect of the wheat bran supplement alone
was detected.
The possibility that drug interactions with dietary fibre or other bulk-
forming herbal drugs may interfere with the gastrointestinal absorption of
orthodox drugs has become of more concern and there have been a number of
pivotal publications. For example, PERLMAN (1990) has reported a case which
raises the possibility of an interaction between lithium salts and ispaghula husk
(psyllium hydrophilic mucilloid), a bulk-forming laxative.
The patient had a long history of psychiatric problems and was treated
with lithium carbonate and then lithium citrate. She also received ispaghula
husk (one teaspoonful in water twice daily). Despite increasing her lithium
dosage, her blood lithium concentrations fell to below acceptable levels.
Ispaghula husk was discontinued and her blood lithium concentrations in-
creased.
RICHTER et al. (1991) have reported that concomitant fibre intake (pectin
15 mg/day or oat bran 50-100 g/day) may decrease the absorption of the lipid-
334 P.A.G.M. DE SMET and P.F. D'ARCY

lowering agent lovastatin. Their patients showed greatly increased low-density

lipoprotein (LDL) cholesterol levels when these agents were introduced into
a lipid-lowering diet plus lovastatin (80mg/day) regimen. LDL cholesterol
levels normalised when the pectin or oat bran was stopped.
STEWART (1992) has reported that excessive dietary fibre may reduce the
efficacy of tricyclic antidepressants. Three of her patients with recurrent maj or
depression, who had been successfully treated with intermittent tricyclics in
the past, became refractory to therapy after beginning to ingest a high-fibre
diet. Serum antidepressant levels were lower than those previously achieved.

2. Gnar Gnm
TODD et al. (1990) have reviewed the pharmacological properties of guar gum.
Administration of up to 30 g/day guar gum to diabetic patients does not seem
to alter mineral or electrolyte balance during long-term therapy (BEHALL et al.
1989; McivOR et al. 1985; UUSITUPA et al. 1989), although plasma levels of fat-
soluble vitamins (A and E) have tended to decrease during extended treat-
ment, but not to a clinically important extent (UUSITUPA et al. 1989).
Because of its effect on prolonging gastric retention and because drugs
will diffuse more slowly out of viscous matrices than from solutions, guar gum
may affect the absorption of certain concomitantly administered drugs. Often
only the rate of absorption is affected [e.g., bumetanide, digoxin, paracetamol
(acetaminophen)] and such interactions are of minor clinical significance.
However, the absorption of certain other drugs (e.g., metformin (GIN et al.
1989), phenoxymethylpenicillin (HUUPPONEN et al. 1985) and some formula-
tions of glibenclamide (NEUGEBAUER et al. 1983) may be reduced by a clinically
significant degree.
Guar gum is a component of many slimming preparations and it might be
advisable that women taking oral contraceptives take additional contraceptive
precautions whilst taking guar gum (ANONYMOUS 1987). The same warning
could be extended to women taking low-dose oestrogen, combined oral con-
traceptives which are on on any slimming diet that contains bulk-forming
agents. Theoretically, they could be at risk of contraceptive failure if the
absorption of the contraceptive is compromised. However, we know of no
clinical study that has assessed the clinical importance of this theoretical
notion, nor any report that has linked slimming diets with contraceptive
failure.

3. Tannins
From the 1960s to the present time, it has been variously suggested and refuted
that among the botanical constituents that might also interfere with the ab-
sorption of some orthodox drugs are tannins in tea or coffee (CARRERA et al.
1973) and phytates (DAVIES 1982). For example, CARRERA et al. (1973) showed
in rat studies that the absorption of vitamin B12 was reduced proportionately to
the oral dose of tannic acid administered. This interaction, it was claimed, was
Drug Interactions with Herbal and Other Non-orthodox Remedies 335

due to the formation of a non-absorbable complex between tannic acid, the

vitamin and glycoproteins in the gut.
There has also been much discussion about the possibility of tea and
coffee comsumption weakening the effect of antipsychotic medication. These
discussions had their foundation in a report by LEVER and HAGUE in 1964 that
phenothiazines can be precipitated in vitro by many fruit juices, milk, tea and
coffee. The mixtures so produced are rendered unpalatable and, since some of
the precipitates do not redissolve in hydrochloric acid, it was suggested that
they might be poorly absorbed in vivo. On this basis it has been variously
suggested that tea and coffee might antagonise the efficacy of antipsychotic
drugs, especially neuroleptic drugs (HIRSCH 1979; KULHANEK et al. 1979).
Support for these views was given by the results of experiments in rats
which suggested that tea and coffee alter the pharmacokinetics of neuroleptic
drugs (KULHANECK and LINDE 1981). Additional support was given by
LASSWELL et al. (1984), who investigated the in vitro interaction of
neuroleptics and tricyclic antidepressants with tea, coffee and gallotannic acid
and they showed that there was significant precipitation with a variety of
agents including several phenothiazines, amitriptyline, haloperidol,
imipramine and loxapine. The strong complex formed between these drugs
and tannins was thought to be the basis of the interaction of these drugs with
tea and coffee. They did not, however, produce any evidence of their own to
suggest that such interactions were of any clinical significance. However, in a
controlled study in 16 female patients in a psychiatric hospital, the effect oftea
and coffee drinking was investigated on steady-state blood levels and clinical
efficacy of chlorpromazine, haloperidol, fluphenazine and trifluoperazine
(BOWEN et al. 1981). Withdrawal of these beverages did not increase the
bioavailability of the drugs studied, nor did they affect the individual variation
in the plasma levels of these drugs. The conclusion was that limitation of tea
and coffee intake in medicated psychiatric patients could not be justified
(BOWEN et al. 1981).
It must be remembered, however, that apart from a tannin-based interac-
tion of caffeinated beverages there is the pharmacodynamic possibility that
there are psychotropic effects of caffeine in psychiatric patients drinking up to
10 cups of coffee per day; these effects of caffeine cannot be neglected (DE
FREITAS and SCHWARTZ 1979).
Ordinary tea is sometimes reputed to affect oral iron absorption and some
researchers have assessed the effects of other herbal teas. HESSELING et al.
(1979) found that freshly prepared rooibos tea (Aspalathus linearis) did not
affect iron absorption in contrast to ordinary tea. More recently, studies by EL-
SHOBAKI et al. (1990) examined the influence of various herbal teas on iron
absorption; they found that ordinary tea inhibited iron absorption, whereas
anise, mint, caraway, cumin, tilia and liquorice promoted the absorption of
iron. It should be noted, however, that other publications dispute that ordinary
tea affects the absorption of pharmacologic doses of oral iron preparations
(KOREN et al. 1982).
336 P.A.G.M. DE SMET and P.P. D'ARCY

II. Effects of Herbal Drugs on Orthodox Drug Pharmacokinetics:

Effects on Elimination

1. Cola Nut
An extensive study by FRASER et al. (1976) of factors affecting antipyrine
metabolism in West African villagers showed that cola nut consumption inhib-
ited antipyrine metabolism and prolonged the antipyrine half-life by 3.Sh. It
was suggested that unidentified constituents of cola nuts competed with
antipyrine for oxidation by the microsomal enzyme system. There were three
other predictors of oxidative capacity: sex, haemoglobin in women and height
in men.
VESELL et al. (1979) failed to show any effect of cola nuts, chewed for
either 14 or 28 consecutive days, on antipyrine disposition in Caucasian males.
Genetic factors may therefore be of importance and although many questions
remain unanswered, the results of the antipyrine studies are interesting
enough to suggest that this type of work be reopened since further research in
this direction and particularly with other herbal remedies would seem to be
warranted.

2. Eucalyptus Species
A number of studies have demonstrated that eucalyptus leaves, the oil and the
major active principle, eucalyptol, can all induce microsomal enzyme activity
in both in vitro and in vivo tests (JORI et al. 1969; SEAWRIGHT et al. 1972).
However, there have not been any recorded interactions between eucalyptus
and orthodox drugs at the clinical level.

3. Grapefruit Juice
There is evidence that grapefruit juice augments the bioavailability of the
antihypertensives felodipine, nifedipine and nitrendipine, and the clearance of
caffeine. These interactions have already been mentioned previously under
"Dietary Fads".

4. Herbal Smoking Preparations

Herbal smoking preparations are sufficiently common in the alternative drug
market (SIEGEL 1976) to suggest the possibility that they, like tobacco smoking,
may contain polycyclic hydrocarbons in their smoke and that they might also
affect the metabolism of orthodox drugs. However, we know of no work done
in this field.

5. Kampo Medicines
Oriental Kampo medicines often contain glycyrrhizin and these have been
reported to influence prednisolone pharmacokinetics. HOMMA et al. (1992a)
Drug Interactions with Herbal and Other Non-orthodox Remedies 337

showed that three Kampo remedies, which also contained Saiko (Bupleuri
radix), and were commonly co-administered with prednisolone in the treat-
ment of asthma, nephrotic syndrome and collagen diseases, had a variable and
different effect on prednisolone pharmacokinetics in healthy sUbjects.
Only one of the three remedies had a steroid-sparing effect due to de-
creased l1,B-hydroxysteroid dehydrogenase (ll-beta-HSD) activity. The other
two either increased or did not change the activity of this enzyme. Interest-
ingly, the authors attributed the enzyme inhibitory effect not to glycyrrhizin,
which was present in the three Kampos, but to magnolol, which was only
contained in the Kampo with a positive effect on prednisolone metabolism.
Follow-up work by HOMMA et al. (1994) was aimed at confirming the
inhibitor of prednisolone metabolism contained in Saiboku-To. In in vitro
experiments they studied the effects of 10 herbal constituents on ll-beta-HSD
and showed that five herbal extracts had inhibitory activity: Glycyrrhiza glabra
> Perillae frutescens > Zizyphus vulgaris> Magnolia officinalis > Scuttellaria
baicalensis. Seven chemical constituents which were identified as the major
urinary products of Saiboku-To in humans were studied; magnolol derived
from M. officina lis showed the most potent inhibition of the enzyme and
although this was less than that of glycyrrhizin, the non-competitive inhibition
mechanism was different from a known competitive mechanism. These results
suggested that magnolol might contribute to the inhibitory effects of Saiboku-
To on prednisolone metabolism through inhibition of ll-beta-HSD.
Studies on ofloxacin in healthy volunteers by HASEGAWA et al. (1994)
found no significant effects of the Kampo medicines Sho-saiko-To (TJ-9),
Rikkunshi-To (TJ-43) and Sairei-To (TJ-1l4) on any estimated bioavailability
parameter.

6. Liquorice
Liquorice, the dried rhizome and roots of Glycyrrhiza glabra, has long been a
popular and traditional ingredient of both herbal and orthodox medicines. It is
also found in confectionary, soft drinks, chewing tobacco and chewing gum.
There is evidence to suggest that liquorice may interfere with the pharmacoki-
netics of orthodox medicines. For example, CHEN et al. (1990) showed that an
intravenous infusion of the active principle, glycyrrhizin (GL), increased the
plasma concentrations of prednisolone and influenced the pharmacokinetics
of prednisolone in man. Interestingly, the main component in plasma is differ-
ent when GL is administered by different routes. After intravenous adminis-
tration the main component in plasma is glycyrrhizin (KATO et al. 1984; CHEN
et al. 1990), while after oral dosage the component is glycyrrhetinic acid (GA),
a metabolic derivative of GL (NAKADA et al. 1986; SHIMADA et al. 1989).
In vitro, the inhibiting effect of GA on the metabolism of corticosteroids
by 5a-,5,B-reductase and ll,B-dehydrogenase is stronger than that of GL. It is
not surprising therefore that in later studies CHEN et al. (1991) showed that
oral administration of GL also modified the pharmacokinetics of both total
338 P.A.G.M. DE SMET and P.F. D' ARCY

and free prednisolone. The area under the curve (AUC) of prednisolone was
significantly increased, the total plasma clearance was significantly reduced,
and the mean residence time was significantly prolonged. However, the vol-
ume of distribution showed no evident change.
These results suggested that the oral administration of GL increases
the plasma prednisolone concentrations and influences its pharmacokinetics
by inhibiting its metabolism, but not affecting its distribution. It has been
suggested that this combination would be advantageous in the treatment
of rheumatoid conditions. The basis of the interaction has been suggested
as inhibition of the metabolism of prednisolone by microsomal enzymes;
inhibition of urine clearance and interference with plasma protein binding
were both discounted. There is direct evidence that GL and GA can inhibit
corticosteroid 5a- and 5p-reductase and IIp-dehydrogenase activities in
rat liver and kidney in vitro and in vivo (TAMURA et al. 1979; MONDER et al.
1989).

III. Effects of Orthodox Drugs on Herbal Drug Pharmacokinetics

Not only may herbal drugs affect the pharmacokinetics of orthodox drugs, but
in reverse orthodox drugs may influence the pharmacokinetics of herbal drugs.
Some illustrative examples are as follows:

1. Caffeine-Containing Herbs
Certain antibacterial 4-quinolones and fluoroquinolones (ciprofloxacin,
enoxacin, pipemidic acid and temafloxacin) inhibit the hepatic metabolism of
caffeine, increase its elimination half-life and decrease its clearance (CARBO et
al. 1989; HARDER et al. 1989; HEALY et al. 1989; MAHR et al. 1992). Users of
caffeine-containing beverages and herbals should therefore be advised that
they have an increased risk of adverse effects (e.g., tremor, tachycardia, insom-
nia, CNS excitation) when they are taking such quinolones. The most impor-
tant herbal remedies which contain substantial amounts of caffeine are
derived from Cola, !lex and Paullinia species (DE SMET 1989).

2. Sparteine-Containing Herb
Sparteine is a quinolizidine alkaloid from Cytisus scoparius which was recently
found in a herbal slimming remedy on the United Kingdom market. Substan-
tial doses of this preparation in slow metabolisers could be expected to
be associated with many adverse reactions including circulatory collapse
(GALLOWAY et al. 1992). The antiarrhythmic agent quinidine is a potent inhibi-
tor of the oxidative metabolism of sparteine (SCHELLENS et al. 1991) and a
similar effect has been observed with haloperidol (GRAM et al. 1989) and with
moclobemide (GRAM et al. 1993).
In addition, CREWE et al. (1992) conducted in vitro experiments in which
several antidepressants (tricyclic or selective serotonin reuptake inhibitors)
Drug Interactions with Herbal and Other Non-orthodox Remedies 339

inhibited human liver microsomal P4502D6 (CYP2D6) activity, which re-

sulted in a reduced oxidative conversion of sparteine to dehydrosparteine.

3. Teucrium chamaedrys
LOEPER et al. (1994) evaluated the molecular mechanism of the hepatotoxicity
which had been observed in numerous users of a French preparation contain-
ing Teucrium chamaedrys. A dose-dependent increase in serum alanine ami-
notransferase (ALT) activity in mice was observed after the intragastric
administration of the lyophilisate of a tea prepared from blooming aerial parts.
Hepatotoxicity could also be produced by intragastric administration of
0.125 g/kg of an enriched fraction, which contained the same level of furano
neo-clerodane diterpenoids as 1.25 g/kg of the lyophilisate. The increase
in serum ALT activity could be enhanced by inducers of the 3A family
(troleandomycin). Toxicity was also increased by pretreatment with phorone
(a depletor of hepatic glutathione), whereas it could be attenuated by inducers
of microsomal epoxide hydrolase (such as clofibrate). These findings suggest
that the hepatotoxicity of T. chamaedrys resides in one or more reactive
metabolites of its furanoditerpenoids and that orthodox drugs may influence
their formation.

IV. Pharmacodynamic Interactions Between Herbal Drugs

and Orthodox Drugs
It is not possible in a chapter of this size to be comprehensive in relating all the
pharmacodynamic interactions involving herbal remedies, it is only possible to
be illustrative. For convenience therefore this present section has been divided
into two categories:
Herbal drugs with well-known constituents, which are also used in ortho-
dox medicine. Many of these are already well known and appreciated and
therefore will be presented in brief detail in tabular form (Table 2). It should
be noted that familiar botanical constituents with potent pharmacological
effects may sometimes occur in very unfamiliar plants. For example, the
traditional Chinese medicine Zangqie, which is derived from Anisodus
tanguticus, yields the same toxic tropane alkaloids as Atropa belladonna and
Datura species (CHANG and BUT 1987).
Herbal drugs with less well-known constituents, which will be discussed in
some more detail here.
Herbal drugs, which do not have generally known potent constituents but
may nevertheless be capable of pharmacodynamic interactions with orthodox
drugs, include the following:

1. Betel Nut (Areca catechu)

Two chronic schizophrenic patients who were maintained on depot
neuroleptics developed serious extrapyramidal symptoms after a period of
340 P.A.G.M. DE SMET and P.F. D'ARCY

Table 2. Herbal drugs with well-known constituents and their interactions. (DE SMET
1992b; D' ARCY 1993)

Herbal drug Interactions

Adonis vernalis Due to the presence of cardioactive glycosides

Convallaria majalis digitalis-like effect and potentiation of toxicity are
Digitalis species possible.
Nerium oleander Although Crataegus extracts do not contain digitalis-
Strophanthus species like glycosides, potentiation of digitalis activity has
Urginea maritima been reported to occur in guinea pigs. The
Xysmalobium undulatum evidential force of this study has been challenged.
Atropa belladonna All these plants contain anticholinergic, tropane
Datura stramonium alkaloids such as hyoscyamine and/or scopolamine,
Hyoscyamus niger which can potentiate the effects of synthetic drugs
Mandragora officinarum with similar pharmacological activity (e.g.,
Scopalia carniolica antidepressants, antihistaminics, antispasmodics).
Ephedra species The tertiary alkaloid ephedrine could interfere with
conventional antihypertensive therapy.
Papaver somniferum The opium alkaloids of the oriental poppy will
interact and potentiate the effects of analgesics and
other CNS depressants.
Pilocarpus pennatifolius The alkaloid pilocarpine has potent cholinergic
effects which may antagonise orthodox treatment
with anti-asthma and anticholinergic agents.
Rauwolfia serpentina The reserpine-rescinnamide alkaloids of this root
will interact and potentiate the effects of
antihypertensives, psychotropics and CNS
depressant drugs.

heavy betel nut chewing. The mechanism for this effect was suggested as
antagonism of the anticholinergic agent procyclidine, by the active alkaloid
ingredient of the betel, arecoline (DEAHL 1989).
TAYLOR et al. (1992) have included betel nut chewing among the factors
that dispose to asthma severity and unsatisfactory control by orthodox medi-
cines in Asians residing in the United Kingdom. It was suggested that
arecoline, or another alkaloid in betel nut, for example, guvacoline, may have
a cholinergic bronchoconstrictor effect.

2. Garlic (Allium sativum)

Since garlic can reduce human platelet aggregation in vitro (BORDIA et al.
1977), it seems difficult to exclude the possibility of untoward circumstances in
patients taking oral anticoagulants. SUNTER (1991) reported seeing two cases
of increased international normalised ratios in patients previously stabilised
on warfarin. The altered anticoagulation picture was attributed to the inges-
tion of garlic products, since there had been no other changes to medication or
habits in either case. One patient had started taking garlic pearls, the other
garlic tablets and in both cases clotting times were roughly doubled. However,
Drug Interactions with Herbal and Other Non-orthodox Remedies 341

an earlier review of the literature (ROCKY MOUNTAIN DRUG CONSULTATION

CENTER 1986) failed to disclose any detailed and well-documented reports of
an interaction between warfarin and garlic and no firm data on prolongation of
prothrombin times by garlic have been published. Nonetheless, SERLIN and
BRECKENRIDGE (1983) have warned that drugs, such as the agents in garlic,
which cause alteration of platelet function, may potentiate warfarin's action
even though the prothrombin time remains unchanged. The present situation
about a possible interaction is unclear and this possibility should be explored
in more detail.

3. Karela (Momordica charantia)

Karela is an oriental folk remedy. It is well established that oral preparations
of the karela fruit have hypoglycaemic activity in a majority of non-insulin-
dependent diabetic patients (LEATHERDALE et al. 1981) and that interference
with conventional treatment by diet and chlorpropamide has been observed
(ASLAM and STOCKLEY 1979). It has also been reported that a subcutaneously
injected principle obtained from the fruit may have a hypoglycaemic effect in
insulin-dependent diabetics (KHANNA et al. 1981).

4. Liquorice (Glycy"hiza glabra)

Apart from its own intrinsic properties of flavouring and sweetening, liquorice
has mild anti-inflammatory and mineralocorticoid properties and may cause
sodium and water retention and hypokalaemia. Control of hypertension may
be difficult if patients taking orthodox antihypertensive drugs also have a
continued high intake of liquorice-containing products (GRIFFIN 1979).
Liquorice may also have adverse effects on diabetic patients maintained on
insulin; glycyrrhizin was associated, not surprisingly because of its mineralo-
corticoid-like effects, with hypokalaemia, and sodium retention (FUJIWARA
et al. 1983).

5. Picrorhiza ku"oa
According to BEDI et al. (1989), the rhizomes of this plant species, which are
used in Ayurvedic medicine, might potentiate the photo chemotherapeutic
effects of methoxsalen in human patients with vitiligo.

v. Multiple or Unclarified Interactions Between Herbal Drugs

and Orthodox Drugs
Medically interesting examples of multiple or unclarified interactions between
herbal drugs and orthodox medications include:

1. Anthranoid Laxatives
An early German report suggests that quinidine serum levels might be re-
duced by the concurrent use of herbal anthranoid laxative preparations, such
342 P.A.G.M. DE SMET and P.F. D'ARCY

as Liquedepur of the Natterman company (GUCKENBIEHL et al. 1976). No

further information is available.

2. Berberine
Berberine is an alkaloid derived from the roots and bark of the plant Berberis
aristata (barberry bush); extracts of this plant have been used in antidiarrhoeal
medication in Ayurvedic medicine in India and in the traditional medicine of
China for the past 3000 years. A Burmese study on the clinical effects of
berberine in acute watery diarrhoea, namely the reputed antisecretory and
vibriostatic effect, showed that berberine alone did not benefit the duration of
diarrhoea, frequency of stools and fluid requirements for rehydration, nor did
it produce a noticeable antisecretory effect. Clinically, patients with cholera
given tetracycline plus berberine were more ill, suffered longer from diarrhoea
and required larger volumes of intravenous fluid than did those given tetracy-
cline alone (KHIN-MAUNG-U et al. 1985).

3. Ginseng (Pamax ginseng)

The concurrent use of ginseng and the MAOI phenelzine has been associated
with adverse effects in two patients (SHADER and GREENBLATT 1985; JONES and
RUNIKIS 1987). However, commercial ginseng preparations are not always
derived from Pamax ginseng and it is therefore difficult to incriminate this
official source plant in the interaction.

4. Piperine
There are several studies to show that piperine, a major alkaloid of Piper
longum and P. nigrum, both of which occur in Ayurvedic formulations, can
enhance the bioavailability of orthodox drugs such as phenytoin, propranolol,
rifampicin, sulphadiazine, tetracycline and theophylline. Among the suggested
mechanisms are promotion of gastrointestinal absorption, inhibition of drug
metabolism and a combination of these two (ATAL et al. 1981, 1985; BANo et
al. 1987, 1991; BHAT and CHANDRASEKHARA 1987; JOHRI and ZUTSHI 1992).

5. "Shankhapusphi"
This is an Ayurvedic non-alcoholic syrup which is prepared from six herbs:
Centella asiatica, Convolvulus pluricaulis, Nardostachys jatamansi, Nepeta
elliptica, Nepeta hindostana and Onosma bracteatum. DANDEKAR et al. (1992)
have reported two epileptic patients taking phenytoin who experienced an
unexpected loss of seizure control and a reduction in plasma phenytoin levels
when they took this herbal preparation. A follow-up study in rats to investi-
gate this possible interaction showed that multidose co-administration (but
not single dose) reduced not only the antiepileptic activity of phenytoin but
also lowered plasma phenytoin levels. Shankhapusphi itself showed significant
Drug Interactions with Herbal and Other Non-orthodox Remedies 343

antiepileptic activity (electroshock seizure prevention) when compared with

placebo.

6. Yohimbine
Yohimbine is a toxic alkaloid from Pausinystalia yohimbe. Preparations pro-
viding pharmacologically relevant doses of yohimbine can occasionally be
encountered on the Western health food market, even though this is a toxic
alkaloid which is not sufficiently safe to be freely available for uncontrolled
use (DE SMET and SMEETS 1994). When given to healthy volunteers, 15-20mg
p.o. yohimbine is usually needed to increase blood pressure and to induce
anxiety, but in patients on tricyclic antidepressants hypertension may already
occur at 4mg t.i.d. (LACOMBLEZ et al. 1989). The toxicity of yohimbine can also
be enhanced by other drugs, such as chlorpromazine, while it is attenuated by
amobarbital or reserpine (INGRAM 1962).
Remarkably, yohimbine may be protrayed in off-label advertising of
yohimbe health food preparations as a peripheral vasodilator, which can po-
tentiate other blood pressure-lowering agents. In reality, however, the a-
adrenoreceptor antagonist properties of the alkaloid will reverse the effects of
clonidine and similar antihypertensives (DE SMET and SMEETS 1994).

VI. Interactions Between Different Herbal Drugs

Herbal remedies are often mixtures, which raises the possibility of interactions
between different herbal ingredients. Thus constituents of different botanical
sources may form complexes with each other. An example of this is the
alkaloid berberine, which combines with the Glycyrrhiza compound
glycyrrhizin to form a complex with modified biopharmaceutical properties.
The alleged basis of the complex is ionic interaction between glucuronic acid
carboxyls and the quaternary centre of berberine (NOGUCHI 1978).
A further example of complex formation in Chinese herbal medicine is
that the Scutellaria constituent, baicalin, a flavonoid glucuronide, is also
complexed with berberine. Glycyrrhetinic acid is also complexed with the
flavonoid glycoside rutin (NOLAN and BRAIN 1985). However, there is no good
evidence to suggest that such complexation compromises the clinical perfor-
mance of these compounds (HOMMA et al. 1992a,b, 1993).
A good South American example of pharmacokinetic interactions be-
tween different herbal ingredients is the native custom to prepare a hallucino-
genic drink called "ayahuasca" from Banisteriopsis vines and Pyschotria
leaves. The former yield beta-carboline alkaloids, but these alkaloids are only
hallucinogenic in high doses and it is doubtful that "ayahuasca" contains
sufficient concentrations of these alkaloids to be hallucinogenic. However,
the Banisteriopsis alkaloids are potent reversible MAOIs and selectively
inhibit MAO-A. The addition of Pyschotria leaves provides the alkaloid
344 P.A.G.M. DE SMET and P.F. D'ARCY

dimethyltryptamine, which is hallucinogenic in low parenteral doses. The oral

dimethyltryptamine would be inactivated by MAO-A but it is thought that
the presence of the Banisteriopsis beta-carbolines in the drink prevents this
untimely degradation (DE SMET 1985; Orr 1994).
One of the many Oriental remedies with multiple herbal ingredients is
"Sairei-to", which is actually a combination of two other multiple prepara-
tions, namely "Sho-saiko-to" and "Gorei-san".
"Sho-saiko-to" consists of seven herbs (Bupleuri radix, Pinelliae tuber,
Scutellariae radix, Ziziphy fructus, Ginseng radix, Glycyrrhizae radix and
Zingiberis rhizoma). "Gorei-san" contains five other herbs (Alismatis
rhizoma, Atractylodis lancea rhizoma, Polyporus, Hoelen and Cinnamomi
cortex) (KIMURA et al. 1990; TSUMURA 1991).
Although the pharmacological activity of "Sairei-to" seems to reside pri-
marily in its "Sho-saiko-to" component, there is some evidence to suggest that
"Gorei-san" is not pharmacologically inert. Studies in ethanol-treated mice
suggest that "Gorei-san" may have a protective effect against certain electro-
lyte deficiencies and may be capable of adjusting fluid and electrolyte metabo-
lism (YONEYAMA et al. 1990). In addition, the deleterious influence of
prednisolone on collagen synthesis in the mouse skin in vivo can be prevented
by "Sairei-to" but not by "Sho-saiko-to" (HANAWA et al. 1987). "Sairei-to"
shows renoprotective activity in rats with aminonucleoside nephrosis
(JOARDER et al. 1991; ITO 1993) and gentamicin-induced nephrotoxicity (OHNO
et al. 1993).

H. Comment
The use of natural (herbal) and other non-orthodox medicines is a persistent
aspect of present-day health care and Europeans alone are thought to spend
the equivalent of US $500-600 million/year on natural remedies and food
supplements. Many consumers believe that naturalness is a guarantee of harm-
lessness and have no qualms, when necessary, in taking their own prescribed
conventional medicine as well. In the United Kingdom and Netherlands
and elsewhere in Europe many immigrant races have their own traditional
medicine practices, which they frequently combine with orthodox medical
care. Generally too little is known about the consequences of such combina-
tions, although the clinical reports of interactions that infrequently appear in
the medical and pharmaceutical press suggest that many more interactions
may be occurring that are not realised as such and are not reported in the
literature.
It should be realised that the data presented in this review did not always
come from studies intended and designed to evaluate adverse interactions
between orthodox drugs and herbal drugs. Some data were merely obtained as
a "spin-off" from pharmacological studies, in which the most useful probe
happened to be a herbal constituent (sparteine, coumarin) or which were
Drug Interactions with Herbal and Other Non-orthodox Remedies 345

undertaken to evaluate the active compound in recreational beverages

(caffeine ).
It is an uncertain state and one that should be actively researched, but it is
complicated at present by lack of knowledge about what many of the herbal
medicines contain. It must be remembered that some herbal drugs may be
contaminated or adulterated with undeclared pesticides, toxic metals, botani-
cals, animal substances and/or orthodox drugs which could lead to additional
and unexpected adverse drug reactions and interactions (GOLDMAN and
MYERSON 1991; JOSEPH et al. 1991; DE SMET 1992a; CAPOBIANCO et al. 1993).
Illustrative is the recent outbreak of food poisoning associated with cucumber
in Dublin (STINSON et al. 1993); this was subsequently traced to be due to the
inappropriate use of "Aldicarb", an anticholinesterase pesticide, by One cu-
cumber producer.
It must be further recalled that quality control of herbals is sometimes
non-existent as indeed may be the quality of the product. However, it is
impossible and unjustified to put all herbal remedies into the same box! There
are reliable preparations On the market in terms of safety and quality and
perhaps the best approach to improve the current uncertain situation is to
promote the good Ones and contrast them with the bad Ones (see Table 1).
Aspects of quality have been highlighted in the "Note for Guidance on the
Quality of Herbal Remedies" by the influential Committee for Proprietary
Medicinal Products of the European Community, which has emphasised the
need to control the purity of herbal remedies (EUROPEAN COMMUNITY 1989).
They must be tested for microbiological quality and for residues of pesticides
and fumigation agents, radioactivity, toxic metals and likely contaminants and
adulterants. Their quality control must be improved and their content of
orthodox synthetic drugs must be established. As to how this is to be achieved
is a matter for national drug regulatory authorities since it is our understand-
ing that the CPMP does not strive for a central role in the regulation of herbal
drugs. Yet a general tightening up of standards is urgently needed if the
alternatively medicated public is to be protected from medication hazard.
Our personal view is that the manufacture and supply of herbal medicines
should be licensed and that the "license escape" category of food supplements
can no longer be justified. If a non-orthodox remedy is recognised and used as
a medicine by the public, and if a dose is specified, then that product should be
regarded as a medicine and not a food supplement, and thus is must conform
with licensing requirements. In contrast to new synthetic medicines (which
should always comply with the regular requirements with respect to efficacy,
safety and purity, we are prepared to accept a separate category for herbal
drugs, for which the requirement of conventional proof of "efficacy" could be
waived. This approach has the disadvantage that it introduces double stan-
dards. However, the present situation is that society wants to use herbal drugs
anyway, and there are many herbal drugs being used that are not controlled at
all. Therefore we prefer to focus On safety and purity (two requirements which
should not be waived lightly) and to ensure thereby the quality and relative
346 P.A.G.M. DE SMET and P.F. D'ARCY

safety of the available products. This approach can only work, however, if it is
supplemented by active herbal pharmacovigilance (DE SMET 1993b), which
should look not only for unknown adverse reactions but also for new adverse
drug reactions (DE SMET 1995).

References
Adams JF, Clarke JS, Ireland JT, Kesson CM, Watson WS (1983) Malabsorption of
vitamin B12 and intrinsic factor secretion during biguanide therapy. Diabetologia
24:16-18
Alvares AP (1984) Environmental influences on drug biotransformations in humans.
World Rev Nutr Diet 43:45-59
Anderson KE (1988) Influences of diet and nutrition on clinical pharmacokinetics. Clin
Pharmacokinet 14:325-346
Anonymous (1987) Guar gum of help to diabetics. Drug Ther Bull 25:65-66
Aslam M, Stockley IH (1979) Interactions between curry ingredient (Karela) and drug
(chlorpropamide). Lancet 1:607
Atal CK, Zutshi U, Rao PG (1981) Scientific evidence on the role of Ayurvedic herbals
on bioavailability of drugs. J Ethnopharmacol 4:229-232
Atal CK, Dubey RK, Singh J (1985) Biochemical basis of enhanced drug bioavailability
by piperine: evidence that piperine is a potent inhibitor of drug metabolism.
J Pharmacol Exp Ther 232:258-262
Bailey DG, Spence JD, Edgar B, Bayliff CD, Arnold JMO (1989) Ethanol enhances
the hemodynamic effects of felodipine. Clin Invest Med 12:357-362
Bailey DG, Spence JD, Munoz C, Arnold JMO (1991) Interaction of citrus juices with
felodipine and nifedipine. Lancet 337:268-269
Bailey DG, Arnold JM, Spence JD (1994) Grapefruit juice and drugs. How significant
is the interaction? Clin Pharmacokinet 26:91-98
Bano G, Raina RK, Sharma DB (1986) Pharmacokinetics of carbamazepine in protein
energy malnutrition. Pharmacology 32:232-236
Bano G, Amla V, Raina RK, Zutshi U, Chopra CL (1987) The effect of piperine on
pharmacokinetics of phenytoin in healthy volunteers. Planta Med 53:568-569
Bano G, Raina RK, Zutshi U, BdiKL, Johri RK, Sharnma SC (1991) Effect of piperine
on bioavailability and pharmacokinetics of propranolol and theophylline in
healthy volunteers. Eur J Clin PharmacoI41:615-617
Barry MG, Macmathuna P, Younger K, Keeling PWN, Feely J (1991) Effect of zinc
supplementation on oxidative drug metabolism in patients with hepatic cirrhosis.
Br J Clin Pharmacol 31:488-491
Bedi KL, Zutshi U, Chopra CL, Amia V (1989) Picrorhiza kurroa, an Ayurvedic herb,
may potentiate photochemothereapy in vitiligo. J Ethnopharmacol 27:347-352
Behall KM, Scholfield DJ, McIvor ME, Van Duyn M, Leo TA et al. (1989) Effects of
guar gum on mineral balances in NIDDM adults. Diabetes Care 12:357-364
Bhat BG, Chandrasekhara N (1987) Interaction of piperine with rat liver microsomes.
Toxicology 44:91-98
Bordia AK, Joshi HK, Sanadhya YK, Bhu N (1977) Effect of essential oil of garlic on
serum fibrinolytic activity in patients with coronary artery disease. Ather-
oscelerosis 28:155-159
Bowen S, Taylor KM, Gibb IAMcL (1981) Effect of coffee and tea on blood levels and
efficacy of antipsychotic drugs. Lancet 1:1217-1218
Braybrooks MP, Barry BW, Abbs ET (1975) The effect of mucin on the bioavailability
of tetracycline from the gastrointestinal tract in vivo, in vitro correlation. J Pharm
Pharmacol 27:508-515
Brown DD, Juhl RP, Warner SL (1977) Decreased bioavailability of digoxin produced
by dietary fiber and cholestyramine. Am J Cardiol 39:297
Brown GR, Greenwood JK (1987) Megavitamin toxicity. Can Pharm J 120:80-87
Drug Interactions with Herbal and Other Non-orthodox Remedies 347

Callaghan TS, Hadden DR, Tomkin GH (1980) Megaloblastic anaemia due to vitamin
B12 absorption associated with long-term metformin treatment. Br Med J 280:
1214-1215
CaIne DB, Sandler M (1970) L-Dopa and Parkinsonism. Nature 226:21-24
Campbell NRC, Hasinoff BB (1991) Iron supplements: a common cause of drug
interactions. Br J Clin Pharmacol 31:251-255
Capobianco DJ, Brazis PW, Fox TP (1993) Proximal-muscle weakness induced by
herbs. N Engl J Med 329:1430
Carbo M, Segura J, De la Torre R, Badenas JM, Cami J (1989) Effect of quinolones on
caffeine disposition. Clin Pharmacol Ther 45:234-240
Carrera G, Mitjavila S, Derache R (1973) Effect de l'acide tannique sur l'absorption de
la vitamine B12 chez Ie rat. C R Acad Paris [D] 276:239-242
Chang H-M, But PPH (eds) (1987) Pharmacology and applications of Chinese materia
medica. World Scientific, Singapore
Chen M-F, Shimada F, Kato H, Yano S, Kanaoka M (1990) Effect of glycyrrhizin on
the pharmacokinetics of prednisolone following low dosage of prednisolone
hemisuccinate. Endocrinol Jpn 37:331-341
Chen M-F, Shimada F, Kato H, Yano S, Kanaoka M (1991) Effect of oral administra-
tion of glycyrrhizin on the pharmacokinetics of prednisolone. Endocrinol Jpn
38:167-175
Chin TF, Lach JL (1975) Drug diffusion and bioavailability: tetracycline metallic
chelation. Am J Hosp Pharm 32:625-629
Cotzias GC (1969) Metabolic modification of some neurological disorders. JAMA
210:1255-1262
Crewe HK, Lennard MS, Tucker GT, Woods FR, Haddock RE (1992) The effect of
selective serotonoin re-uptake inhibitors on cytochrome P4502D6 (CYP2D6) ac-
tivity in human liver microsomes. Br J Clin Pharmacol 34:262-265
Dandekar UP, Chandra RS, Dalvi SS, Joshi MV, Gokhale PC, Sharma MV, Shah PU,
Kshirsagar NA (1992) Analysis of a clinically important interaction between
phenytoin and Shankhapushpi, an Ayurvedic preparation. J Ethnopharmacol
35:285-288
D'Arcy PF (1991) Adverse reactions and interactions with herbal medicines. I. Ad-
verse reactions. Adverse Drug React Toxicol Rev 10:189-208
D' Arcy PF (1993) Adverse reactions and interactions with herbal medicines. II. Drug
interactions. Adverse Drug React Toxicol Rev 12:147-162
Davies NT (1982) Effects of phytic acid on mineral availability. In: Vahouny GV,
Krtichevshky D (eds) Dietary fiber in health and disease. Plenum, New York, pp
105-116
De Freitas B, Schwartz G (1979) Effects of caffeine in chronic psychiatric patients. Am
J Psychiatry 136:1337-1338
De Smet P AGM (1985) Ritual enemas and snuffs in the Americas. Foris, Dordrecht
(Latin America studies no 33, Centre de Estudios y Documentacion
Latinoamericanos)
De Smet PAGM (1989) Drugs used in non-orthodox medicine. In: Dukes MNG,
Beeley L (eds) Side effects of drugs annual 13. Elsevier, Amsterdam, pp 442-
473
De Smet PAGM (1991) Drugs used in non-orthodox medicine. In: Dukes NMG,
Aronson JK (eds) Side effects of drugs annual 15. Elsevier, Amsterdam, pp 514-
531
De Smet PAGM (1992a) Toxicological outlook on the quality assurance of herbal
remedies. In: De Smet PAGM, Keller K, Hansel R, Chandler RF (eds) Adverse
effects of herbal drugs, voll. Springer, Berlin Heidelberg New York, pp 1-72
De Smet PAGM (1992b) Drugs used in non-orthodox medicine. In: Dukes MNG (ed)
Side effects of drugs. Elsevier, Amsterdam, pp 1209-1232
De Smet PAGM (1993a) Legislatory outlook on the safety of herbal medicines. In: De
Smet PAGM, Hansel R, Kellere K, Chandler RF (eds) Adverse effects of herbal
drugs, vol 2. Springer, Berlin Heidelberg New York, pp 1-90
348 P.A.G.M. DE SMET and P.F. D'ARCY

De Smet PAGM (1993b) An introduction to herbal pharmacoepidemiology. J

EthnopharmacoI38:197-208
De Smet P AGM (1995) Should herbal remedies be licensed as medicines? (invited
editorial). Br Med J 310:1023-1024
De Smet PAGM, Smeets OSNM (1994) Potential risks of health food products contain-
ing yohimbine extracts. Br Med J 309:958
Deahl M (1989) Betel nut-induced extrapyramidal syndrome: an unusual drug interac-
tion. Mov Disord 4:330-333
Dent CE, Richens A, Rowe DJF, Stamp TCB (1970) Osteomalacia with long-term
anticonvulsant therapy in epilepsy. Br Med J 4:69-72
Duvoisin RC, Yahr MD, Cote LD (1969) Pyridoxine reversal of L-dopa effects in
Parkinsonism. Trans Am Neurol Assoc 94:81-84
Edgar B, Bailey D, Beregstrand R, Johnsson G, Regardh CG (1992) Acute effects of
drinking grapefruit juice on the pharmacokinetics and dynamics of felodipine -
and its potential clinical relevance. Eur J Clin PharmacoI42:313-317
EI-Shobaki FA, Saleh ZA, Saleh N (1990) The effect of some beverage extracts on
intestinal iron absorption. Z Ernahrungswiss 29:264-269
European Community (1989) Quality of herbal remedies. The rules governing medici-
nal products in the European Community, vol 3. Guidelines on the quality, safety
and efficacy of medicinal products for human use. Office for the Official Publica-
tions of the European Community, Luxembourg, pp 31-37
Evans CDH, Lacey JH (1986) Toxicity of vitamins: complications of a health move-
ment. Br Med J 292:509-510
Fels JP, Neccairi J, Toussain P, Debry G, Luyckx A, Scheen A (1989) Effect of food
intake on kinetics and bioavailability of (4-chlorophenyl) thiomethylene bis-
phosphonic acid (abstract). Calcif Tissue Int 44 [Suppl]:S-104
Fleisch H (1987) Experimental basis for the use of bisphosphonates in Paget's disease
of bone. Clin Orthop 217:72-78
Floyd RA, Greenberg WM, Caldwell C (1977) In vitro interaction between digoxin and
bran. 12th Annual ASHP Midyear Clinical Meeting, Dec 6, Atlanta
Fraser HS, Bulpitt CJ, Kahn C, Mould G, Mucklow JC, Dollery CT (1976) Factors
affecting antipyrine metabolism in West African villagers. Clin Pharmacol Ther
22:369-376
Fuhr U, Klittich K, Staib AH (1993) Inhibitory effectd of grapefruit juice and its bitter
principle, naringenin on CYP1A2 dependent metabolism of caffeine in man. Br J
Clin PharmacoI35:431-436
Fujiwara Y, Kikkawa R, Nakata K, Kitamura E, Takama T, Shigeta Y (1983)
Hypokalaemia and sodium retention in patients with diabetes and chronic hepati-
tis receiving insulin and glycyrrhizin. Endocrinol Jpn 30:243-249
Fushimi R, Tachi J, Amino N, Miyai K (1989) Chinese medicine interfering with
digoxin immunoassay. Lancet 1:339
Fushimi R, Koh T, Tyama S, Yasuhara M, Tachi J, Kohda K, Amino N, Miyai K (1990)
Digoxin-like immunoreactivity in Chinese medicine. Ther Drug Monit 12:242-245
Gallowary JH, Farmer K, Weeks GR, Marsh ID, Forrest ARW (1992) Potentially
hazardous compound in herbal slimming remedy. Lancet 340:179
Gin H, Orgerie MB, Aubetin J (1989) The influence of guar gum on absorption of
metformin from the gut in healthy volunteers. Horm Metab Res 21:81-83
Goldman JA, Myerson G (1991) Chinese herbal medicine; camouflaged prescription
antiinflammatory drugs, corticosteroids and lead. Arthritis Rheum 34:1207
Gram LF, Debruyne C, Caillard V, Boulenger JP, Lacotte J, Moulin M, Zarifian E
(1989) Substantial rise in sparteine metabolic ratio during haloperidol treatment.
Br J Clin Pharmacol 27:272-275
Gram LF, Brosen K, Danish University Antidepressant Group (1993) Moclobemide
treatment causes a substantial rise in the sparteine metabolic ratio. Br J Clin
Pharmacol 35:649-652
Griffin JP (1979) Drug-induced disorders of mineral metabolism. In: D' Arcy PF, Grif-
fin JP (eds) Iatrogenic diseases, 2nd edn. Oxford University Press, Oxford, pp 226-
238
Drug Interactions with Herbal and Other Non-orthodox Remedies 349

Griffin JP, D'Arcy PF, Speirs CJ (1988) A manual of adverse drug interactions, 4th edn.
Wright, (Butterworths) London
Guckenbiehl W, Gilfrich HJ, Just H (1976) EinfluB von Laxantien und Metoclopramid
auf die Chinidin-Plasmakonzentration wahrend Langzseittherapie bei Patienten
mit HerzrythmusstOrungen. Med Welt 27:1273
Hanawa T, Hirama N, Kosoto H, Ja Hyun S, Ohwada S, Hasegawa R, Haranaka R,
Nakagawa S (1987) Effects of Saiko-zai on collagen metabolism in mice adminis-
tered with glucocorticoid. Nohon Univ J Med 29:197-205
Hansten PD, Horn JR, Koda-Kimble MA, Young L Y (eds) (1993) Anticonvulsant
drug interactions - folic acid (Folvite). Drug Interact Updates Q 13:355-356
Harder S, Fuhr U, Staib AH, Wolff T (1989) Ciprofloxacin-caffeine: a drug interaction
established using in vivo and in vitro investigations. Am J Med 87:89S-91S
Hasegawa T, Yamaki K, Nadai M, Muraoka I, Wang L, Takagi K, Nabeshima T (1994)
Lack of effect of Chinese medicines on bioavailability of ofloxacin in healthy
volunteers. Int J Clin Pharmacol Ther 32:57-61
Heald GE, Poller L (1974) Anticoagulants and treatment for chilblains. Br Med J
2:455
Healy DP, Polk RE, Kanawati L, Rock DT, Mooney ML (1989) Interaction between
oral ciprofloxacin and caffeine in normal volunteers. Antimicrob Agents
Chemother 33:474--478
Helsing E (1992) Vitamins. In: Dukes MNG (ed) Side effect of drugs, 11 edn. Elsevier,
Amsterdam, pp 959-976
Hesseling PB, Klopper JF, van Heerden PD (1979) Die effek van Rooibostee op
ysterabsorpsie. S Afr Med J 55:631-632
Hirsch SR (1979) Precipitation of antipsychotic drug in interactions with coffee or tea.
Lancet 11:1130-1131
Hogan RP (1983) Hemorrhagic diathesis caused by drinking a herbal tea. JAMA
249:2679-2680
Homma M, Oka K, Ikeshima K, Takahashi N, Yamamoto S, Niitsuma T, Hoh H
(1992a) Different effects of three Kampo-remedies which compromise of saiko on
prednisolone pharmacokinetics. 5th World Conference on Clinical Pharmacology
and Therapeutics, July 26-31, Yokohama
Homma M, OkaK, Yamada T, Niitsuma T, Itoh H, Takahashi N (1992b) A strategy for
discovering biologically active compoounds with high probability in traditional
Chinese herb remedies: an application of Saiboku-to in bronchial asthma. Anal
Biochem 203:179-187
Homma M, Oka K, Niitsuma T, Hoh H (1993) Pharmacokinetic evaluation of tradi-
tional Chinese herbal remedies. Lancet 341:1595
Homma M, Oka K, Niitsuma T, Hoh H (1994) A novel 11 beta-hydroxysteroid dehy-
drogenase inhibitor contained in Saiboku-To, a herbal remedy for steeroid-depen-
dent bronchial asthma. J Pharm Pharmacol 46:305-309
Hoppe HA (1975) Angiospermen, 8 edn. Gruyter, Berlin, pp 91-92, 435, 519-520, 700-
701 (Drogenkunde, vol 1)
Huupponen R, Karhuvaara S, Seppala P (1985) Effect of guar gum on glipizide absorp-
tion in man. Eur J Clin PharmacoI28:717-719
Ingram CG (1962) Some pharamcologic actions of yohimbine and chlorpromazine in
man. Clin Pharmacol Ther 3:345-352
Ito K (1993) Effects of Sairei-to on nephrotic syndrome. J Tokyo Womens Med Coll
63:464-468
Joarder ZH, Ogawa T, Yorioka N, Yamakido M (1991) Studies on the effectiveness of
Sairei-to on puromycin aminonucleoside nephrosis in rats. Hiroshima J Med Sci
40:127-135
Johri RK, Zutshi U (1992) An Ayurvedic formulation "Trikatu" and its constituents.
J Ethnopharmacol 37:85-91
Jones BD, Runikis AM (1987) Interaction with ginseng. J Clin Psychopharmacol
7:201-202
Jori A, Bianchetti A, Prestini PE (1969) Effect of essential oils on drug metabolism.
Biochem Pharmacol 18:2081-2085
350 P.A.G.M. DE SMET and P.F. D'ARCY

Joseph AM, Biggs T, Garr M, Singh J, Lederle FA (1991) Stealth steroids. N Engl J
Med 324:62
Karunanithy R, Sumita KP (1991) Undeclared drugs in Chinese antirheumatoid medi-
cine. Int J Pharm Pract 1:117-119
Kato H, Nakanishi K, Kanoaka M (1984) The enzyme immunoassay of glycyrrhizin and
glycyrrhetinic acid and its clinical application. Minophagen Med Rev 15 [Suppl]:3-
11
Kempin SJ (1983) Warfarin resistance caused by broccoli. N Engl J Med 308:1229-1230
Khanna P, Jain SC, Panagariya A, Dixit VP (1981) Hypoglycemic activity of polypep-
tide-P from plant source. J Nat Prod 44:648-655
Khin-Maung-U, Myo-Khin, Nyunt-Nyunt-Wai, Aye-Kyaw, Tin-U (1985) Clinical trial
of berberine in acute watery diarrhoea. Br Med J 291:1601-1605
Kimura K, Nanba S, Tojo A, Matsuoka H, Sugimoto T (1990) Effects of Sairei-to
on the relapse of steroid-dependent nephrotic syndrome. Am J Clin Med 18:45-
50
Koren G, Boichis H, Keren G (1982) Effect of tea on the absorption of pharmacologic
doses of an oral iron preparation. Isr J Med Sci 18:547
Kulhanek F, Linde OK (1981) Coffee and tea influence pharmacokinetics of
antipsychotic drugs. Lancet 11:359-360
Kulhanek F, Linde OK, Meisenberg G (1979) Precipitation of antipsychotic drugs in
interaction with coffee and tea. Lancet 11:1130
Kunin CM, Finland M (1961) Clinical pharmacology of the tetracycline antibiotics. Clin
Pharmacol Ther 2:51-69
Kwan T, Paiusco AD, Kohl L (1992) Digitalis toxicity caused by toad venom. Chest
102:949-950
Lacomblez L, Bensimon G, Isnard F, Diquet B, Lecrubier Y, Puech AJ (1989) Effect
of yohimbine on blood pressure in patients with depression and orthostatic
hypotension induced by clomipramine. Clin Pharmacol Ther 45:241-251
La Piana Simonsen L (1989) Top 200 drugs of 1988. Pharmacy Times 40
Lasswell WL Jr, Weber SS, Wilkins JM (1984) In vitro interaction of neuroleptics and
tricyclic antidepressants with coffee, tea and gallotanic acid. J Pharm Sci 73:1056-
1058
Leatherdale BA, Panesar RK, Sing G, Atkins TW, Bailey CJ, Bignell AH (1981)
Improvement in glucose tolerance due to Momordica charantia (karela). Br Med
J 282:1823-1824
Lever PG, Hague JR (1964) Observations on phenothiazine concentrates and diluting
agents. Am J Psychiatry 120:1000-1002
Lewis WH, Elvin-Lewis MPF (1977) Medical botany. Plants affecting man's health.
Wiley, New York, p 192
Lin C-S, Lin M-C, Chen K-S, Ho C-C, Tsai S-R, Ho C-S, Shieh W-H (1989) A digoxin-
like immunoreactive substance and atrioventricular block induced by a Chinese
medicine "Kyushin". Jpn Circ J 53:1077-1080
Loeper J, Descatoire V, Letteron P, Moulis C, Degott C, Dansette P, Fau D, Pessayre
D (1994) Hepatotoxicity of germander in mice. Gastroenterology 106:464-472
Mahr G, Soergel F, Granneman GR, Kinzig M, Muth P, Patterson K, Fuhr U, Nickel
P, Stephan U (1992) Effects of temafloxacin and ciprofloxacin on the pharmacoki-
netics of caffeine. Clin Pharmacokin 22 [Suppl 1]:90-97
McIvor ME, Cummings CC, Mendeloff AI (1985) Long-term ingestion of guar gum is
not toxic in patients with noninsulin-dependent diabetes mellitus. Am J Clin Nutr
41:891-894
Medical Letter (1971) Vitamin C - were the trials well conducted and are large doses
safe? Med Lett Drugs Ther 13:48
Merkel U, Sigusch H, Hoffmann A (1994) Grapefruit juice inhibits 7-hydroxylation of
coumarin in healthy volunteers. Eur J Clin PharmacoI46:175-177
Monder C, Stewart PM, Lakshmi V, Valentino R, Burt D, Edward CRW (1989)
Licorice inhibits corticosteroid II-dehydrogenase of rat kidney and liver: in vitro
and in vivo studies. Endocrinology 125:1046-1053
Drug Interactions with Herbal and Other Non-orthodox Remedies 351

Nakada T, Kato H, Yano S, Kanaoka M (1986) Pharmacokinetics of glycyrrhizin and

glycyrrhetinic acid in healthy humans with long-term treatment of glycyrrhizin
preparations (1). J Med Pharm Soc WAKAN-YAKU 3:278-279
Neugebauer G, Akpan W, Abshagen U (1983) Interaktion von Guar mit
Glibenclamide und Bezafibrat. Beitr Infusionsther Klin Ernahrung 12:40-47
Neuvonen PJ (1976) Interactions with the absorption of tetracyclines. Drugs 11:45-54
Noguchi M (1978) Studies on the pharmaceutical quality evaluation of crude drug
preparations used in orient medicine "Kampoo". II. Precipitation reaction of
berberine and glycyrrhizin in aqueous solution. Chern Pharm Bull (Tokyo)
26:2624-2629
Nolan JE, Brain KR (1985) Component interaction in Chinese herbal medicines. Acta
Agronom 34 [Suppl):113
Ohno I, Shibasaki T, Nakano H, Matsuda H, Matsumoto H, Misawa T, Ishimoto F,
Sakai 0 (1993) Effects of Sairei-to on gentamicin nephrotoxicity in rats. Arch
ToxicoI67:145-147
Ott J (1994) Ayahuasca analogues. Pangaean entheogens. Natural Products,
Kennewick
Panesar NS (1992) Bufalin and unidentified substance(s) in traditional Chinese medi-
cine cross-react in commercial digoxin assay. Clin Chern 38:2155-2156
Patriarca PA, Kendal AP, Stricof RL, Weber JA, Meissner MK, Dateno B, (1983)
Influenza vaccination and warfarin or theophylline toxicity in nursing-home resi-
dents. N Engl J Med 308:1601-1602
Perlman BB (1990) Interaction between lithium salt and ispaghula husk. Lancet
335:416
Pierpaoli PG (1972) Drug therapy and diet. Drug Intell Clin Pharm 6:89-99
Reissell P, Manninen V (1982) Effect of administration of activated charcoal and fibre
on absorption, excretion and steady state blood levels of digoxin and digitoxin.
Evidence for intestinal secretion of the glycosides. Acta Med Scand 668
[Suppl):88-90
Richter WO, Jacob BG, Schwandi P (1991) Interaction between fibre and lovastatin.
Lancet 338:706
Robinson MF, Thomson CD, Huemmer PK (1985) Effect of a megadose of ascorbic
acid, a meal and orange juice on the absorption of selenium as sodium selenite.
N Z Med J 98:627
Rocky Mountain Drug Consultation Center (1986) Warfarin-garlic interaction.
Micromedex, Denver (Drugdex, vol 61)
Roe DA, Kalkwarf H, Stevens J (1988) Effect of fiber supplements on the apparent
absorption of pharmacological doses of riboflavin. J Am Diet Assoc 88:211-213
Rossander L (1987) Effect of dietary fiber on iron absorption in man. Scand J
Gastroenterol 129 [Suppl):68-72
Schellens JH, Ghabrial H, van der Wart HH, Bakker EN, Wilkinson GR, Breimer DD
(1991) Differential effects of quinidine on the disposition of nifedipine, sparteine,
and mephenytoin in humans. Clin Pharmacol Ther 50:520-528
Seawright AA, Steele DP, Menrath RE (1972) Seasonal variation in hepatic microso-
mal oxidation metabolism in vitro and susceptibility to carbon tetrachloride in a
flock of sheep. Aust Vet J 48:488-494
Serlin MJ, Breckenridge AM (1983) Drug interactions with warfarin. Drugs 25:610-620
Shader RI, Greenblatt DJ (1985) Phenelzine and the dream machine - ramblings and
reflections. J Clin Psychopharmacol 5:65
Shimada F, Chen MF, Kato H, Yano S, Kanaoka M (1989) Pharmacokinetics of
glycyrrhizin and glycyrrhetinic acid in healthy humans with long-term treatment of
glycyrrhizin preparations (II). J Med Pharm Soc WAKAN-YAKU 6:402-403
Siegel RK (1976) Herbal intoxication - psychoactive effects from herbal cigarettes, tea,
and capsules. JAMA 236:473-476
Sokol RJ, Johnson KE, Karrer KE, Narkewicz MR, Smith D, Kam I (1991) Improve-
ment of cyclosporing absorption in children after liver transplantation by means of
water-soluble vitamin E. Lancet 338:212-214
352 P.A.G.M. DE SMET and P.F. D'ARCY: Drug Interactions

Soons PA, Vogels BAPM, Roosemalen CM, Schoemaker HC, Uchida E, Edgar B,
Lundahl J, Cohen AF, Breimer DD (1991) Grapefruit juice and cimetidine inhibit
stereos elective metabolism of nitrendipine 'in humans. Clin Pharmacol Ther
50:394-403
Stewart DE (1992) High-fiber diet and serum tricyclic antidepressant levels. J Clin
Psychopharmacol 12:438-440
Stinson JC, O'Gharabhain F, Adebayo G, Chambers PL, Feely J (1993) Pesticide-
contaminated cucumber. Lancet 341:64
Stockley IH (1984) Warfarin resistance caused by over-the-counter drugs, food supple-
ments and vegetables. Pharm Int 5:165-167
Stockley IH (1991) Drug interactions. A source book of adverse interactions, their
mechanisms, clinical importance and management, 2nd edn. Blackwell, Oxford
Sunter WH (1991) Warfarin and garlic. Pharm J 246:722
Tamura Y, Nishikawa T, Yamada K, Yamamoto M, Kumagai A (1979) Effects of
glycyrrhetinic acid and its derivatives on ~4-5a- and 5f3-reductase in rat liver.
Arzneimittelforschung 26:647-649
Taylor RFH, AI-Jarad N, John LME, Conroy DM, Barnes NC (1992) Betel-nut chew-
ing and asthma. Lancet 339:1134-1136
Teuscher E, Lindequist U (1988) Biogene Gifte. Biologie-Chemie-Pharmakologie.
Academie, Berlin, pp 228-229
Todd P A, Benfield P, Goa KL (1990) Guar gum. A review of its pharmacological
properties and use as a dietary adjunct in hypercholesteraemia. Drugs 39:917-928
Tsumura A (1991) Kampo - how the Japanese updated traditional herbal medicine.
Japan Publications, Tokyo
Udall JA, Krock LB (1968) A modified method of anticoagulant therapy. Curr Ther
Res 10:207-211
Uusitupa M, Siitonen 0, Savolainen K, Silvasti M, Penttila I, Parviainen M (1989)
Metabolic and nutritional effects of long-term use of guar gum in the treatment
of noninsulin-dependent diabetes of poor metabolic control. Am J Clin Nutr
49:345-351
Vesell ES, Shively CA, Passananti GT (1979) Failure of cola nut chewing to alter
antipyrine disposition in normal male subjects from a small town in South Central
Pennsylvania. Clin Pharmacol Ther 26:287-293
Walker FB (1984) Myocardial infarction after diet-induced warfarin resistance. Arch
Intern Med 144:2089-2090
Watson AJM, Pegg M, Green JRB (1984) Enteral feeds may antagonise warfarin. Br
Med J 288:557
Wesley-Hadzija B (1971) A note on the influence of diet in West Africa on urinary pH
and excretion of amphetamine in man. J Pharm Pharmacol 23:366-368
Yoneyama Y, Usukura Y, Arai H, Hasegawa R, Haranaka R, Nakagawa S (1990)
Effects of Sairei-to on glutathione metabolism in alcohol-treated mice. Nihon
Univ J Med 32:379-387
Subject Index

absorption allylisopropylacetamide 164

active 47 2-allyloxy-4-chloro-N-(2-
carrier mediated 29 diethylaminoethyl)-benzamide 28
passive 28 2-allyloxy-3-methylbenzamide 296
window 25 aloe 312
acarbose 227,244,245 alphaxalone 296
accelerated drug absorption alprazolam 79,98,99,288
food interactions causing 95 aluminium 296
acetaminophen 22, 29, 30, 153, 286, aluminum hydroxide 16, 17,26,240
288,308,316,334 Alzheimer's disease 223
acetanilide 286, 299, 312 amantadine 197,220
acetazolimide 296 ambenonium chloride 51, 52, 55, 56
acetohexamide 310 amidopyrine 296
acetone 153 amiloride 188, 193, 199
acetorphan 63, 64 Aminodur 269
acetylator status 299 aminoglutethimide 296
acetylcholine 222 aminoglycosides 223
acetylcysteine 316 p-aminohippuric acid 175
N-acetylcysteine 318 aminophylline 246,316
f3-acetyldigoxin 37 aminopyrine 286, 312
n-acetylprocainamide 188, 189, 191, aminosalicylic acid 312
194,195,196,197,316 p-aminosalicylic acid 8, 14, 17
~-acid glycoprotein 125,130,139 5-aminosalicylic acid 64, 70
acid secretion 47 amiodarone 79,85,162,163,198
acyclovir 181 amitriptyline 286,288,296,297,312,
Adonis vernalis 340 335
adrenal corticosteroids 226 amlodipine 98, 99, 104
adrenaline 3,215,217,265,296 amobarbital 286, 288, 343
adrenochrome pigmentation 265 amocarzine 79,85
Aerolate 269 amoxycillin 31, 185,242
afiatoxine Bl 153 amphetamine 152,216,219,220,221
afiatoxins 153 amphetamine sulphate 271
albumin 125, 285 ampicillin 31,286,317
binding sites 126 angiotensin converting enzyme
albuterol 64 inhibitors 227,228,229,230,289
alclofenac 296 N',N' -anhydro-bis-(f3-hydroxy-
alcohol 160, 161,296 ethyl)biguanide 28
alcohol dehydrogenase 165 aniracetam 63, 64
Aldicarb 345 anisindione 312
aldose reductase inhibitors 244 antacids 14,16,17
alendronate sodium 51, 52 anthranoid laxatives 341
alkaline phosphatase analysis 318 antileprosy treatments 240
allopurinol 162, 163, 178, 306 antipyrine 157,286,288,297,312,336
alloxanthine 306 apronalide 296
354 Subject Index

aqueous channels 29 brompheniramine 102

ara-AC (fazarabine) 245 buflomedil 80, 86
arachidonic acid 155 bufuralol 156, 298
arecoline 340 bumetanide 203, 296, 334
arsenic 327 bupivacine 256, 296
arylamine 153 buprenorphine 296
ascorbic acid 308, 309 busulphan 296
aspirin 14, 130, 186,205, 206, 282, 286, butazolidin 8
296,308,313
enteric coated 14 566C80 84, 94
astemizole 79, 163, 226 882C 54,63
atenolol 52, 55, 56 Caco-2 cells 32
atovaquone 79, 85 caffeine 153, 157, 316, 317, 332, 335,
Atropa belladonna 340 336
atropine 15, 26, 223, 262, 296 cancer chemotherapy 245
attapulgite and pectin 16 captopril 31, 32, 298, 331
avapyrazone 296 carbamazepine 14,21,99,141,160,
ayahuasca 343 223,239,240,246,296
azapropazone 187, 296 carbenicillin 286, 316
azathioprine 162 carbenoxolone 160, 282, 286
azithromycin 52, 56 carbidopa 53, 60
azuresin 312 carbimazole 298
carbon disulphide 164
baclofen 224 carbon tetrachloride 164
baicalin 343 carboplatin 251
bambuterol 98, 99 carbromal 296
barbiturates 160 carcinogen (NNK) 153
Bay-X-1005 80,86 cardizem 99
bemegride 296 carmustine 130, 245, 255
bendrofluazide 202, 203 L-carnosine 31
1,2-benzanthrene 154 cascara 312
benzhexol 14 cefaclor 64, 65, 70, 72, 183
benzoic acid 250 cefadroxil 183
benzol a]pyrene 153 cefamandole 183, 286
3,4-benzpyrene 159 cefatrizine 30
benzylamine 3 cefazedone 183
benzylpenicillin 181, 185 cefazolin 183, 286, 310
berberine 342, 343 cefdinir 14, 19, 64
betel nut 339 cefetamet pivoxil 80,86,87,88,99,104
bethanidine 296 cefmenoxime 183
bicuculline 224 cefmetazole 183
biguanides 296 cefonicid 183
bile salts 20, 28 ceforanide 183
complexation with 21 cefotaxime 183
bile secretion 47 cefoxitin 183,310, 311
binding sites, displacement of cefprozil 52,57,65,71, 72
drugs from 129 cefradine 286
bismuth 242 ceftazidine 183
bisoprolol 98, 99 ceftiputen 52
bleomycin 255 ceftizoxime 183
blood-brain barrier 106 ceftriaxone 134, 182, 183
British National Formulary 238,240, cefuroxime 80,87,183
244 cefuroxime axe til 246
British Thoracic Society 239 celiprolol 33
brofaromine 3, 80, 86, 99 cellulose acetate 253
bromocriptine 99, 104,220 propionate 253
Subject Index 355

cephalexin 14, 31, 56, 183, 194, 296, 316 closapine 219
cephalothin 183, 194, 310, 311 cocaine 216,217,219,220,221
cephradine 29, 31, 182, 183 codeine 296,316
CGP 43371 80,87,88 cola nut 336
CHAP 5 245 colchicine 17, 296, 327
charcoal 14,16,21,313 colestipol 14
chelate formation 19 colistine 296
chinese toad venom 329 Committee for Proprietary Medicinal
chloral hydrate 26, 130, 296 Products of the European
chlorambucil 296 Community 345
chloramphenicol 160,162,296,317 contact lenses, drug interactions with
chlorazepate, dipotassium 256 265
chlorcyclizine 161 Convallaria majalis 340
chlordiazepoxide 26, 138, 263, 286, 288, coumarin 327,332,344
296 creatinine 310, 311, 317
chlorinated hydrocarbons 160 analysis 317
chlormethiazole 264, 286, 296 clearance 190, 202
chlormezanone 296 Crohn's disease 126
2-chloro-2'-deoxyadenosine 52, 56 cyanide 309
chloroform 296 cyclacillin 31
chloroquine 6,226,266,296,312,332 cyclamate sodium 15
chlorothiazide 14,25,203 cyclizine 296
chlorpheniramine 271,296 cyclophosphamide 37, 162, 164,245,
chlorpromazine 14,152,220,223,226, 286, 296
255,284,296,298,335,343 cyclopropane 296
chlorpropamide 206,296,341 cyclosporine 6, 34, 35, 81, 89, 100, 105,
chlortetracycline 14, 19 153,161,162,163,227,246,263,265
chlorthalidone 286 cyclosporine A 263
chlorzoxazone 153,312 CYPIA 152
cholestyramine 14,17,20 CYPIAI 153
cholesterol 155 CYPIA2 153, 157, 334
cholecystokinin 8, 87 CYP2A6 153, 157
choline 222 CYP2A6 v 157
cholinergic receptors 223 CYP2A7 153
synapse 222 CYP2B 152
cicaprost 52 CYP2B6 153
cimetidine 4, 100, 105, 160, 162, 164, CYP2B7 153
176, 188, 189, 190, 191, 192, 193, CYP2C 152
194,240,246,283,286,296,312 CYP2C9 153
cinoxacin 182 CYP2CI0 153
ciprofioxacin 14,27, 52, 57, 61, 160, CYP2C17 153
161,162,338 CYP2C18 153
circadian rhythm 76,77 CYP2C19 153
cisapride 223 CYP2D 152
cisplatin 245,251, 271 CYP2D6 153, 339
citalopram 220,221 CYP2D7P 153
CL 275838 65 CYP2D8P 153
clarithromycin 80, 81, 89, 241, 242 CYP2E 152
clindamycin 14,134 CYP2El 153
clobazam 288, 296 CYP2F 152
clofazimine 240, 246, 313 CYP2Fl 153
clofibrate 160, 179, 296 CYP3A 34, 35, 152, 154
clomipramine 220, 256 CYP3A3 153, 154
clonazepam 240, 296 CYP3A4 49,153,154,157,332
clonidine 216,217,218,296 CYP3A5 153, 157
clorgyline 3 CYP3A7 153
356 Subject Index

cysteine 318 dimenhydrinate 296

cytochrome P448 159 dimethylaniline 152, 165
cytochrome P450 4,34,151,155,164, dimethylsulfoxide 312
287 dimethyltryptamine 344
effects of age and disease on 157 diphenadione 312
genes 152 diphenhydramine 14, 296
monooxygenase 151, 152 dipyrone 297
polymorphism 156 disopyramide 138, 141, 161, 192, 193,
cytostatic drugs 37 217,227,296,298
enantiomers of 192
dactinomycin 245 dithiazinine 313
danazol 81,89,162,163,296 diuretic combinations 237
danthron 312,313 domperidone 220, 296
dapsone 161,240,296,299 dopamine 3,217,220
Datura stramonium 340 dopaminergic synapse 219
daunorubicin 312 dorzolamide 127
debrisoquine 93,153,157,298 doxazosin 5, 59
hydroxylase 156 doxorubicin 245, 255, 312
deferoxamine mesylate 312 doxycycline 15, 65, 286
dehydroepiandrosterone 155 droperidol 296
dehydrofelodipine 35, 36, 90 drug binding, displacement by
dehydronorfedipine 90 endogenous materials 138
delayed absorption, food interactions drug metabolising enzymes 151
causing 63 drug interactions at receptor sites 215
desalkylflurazepam 288 in vitro 249
desipramine 286 synergistic 235
desmethyldiazepam 286, 288 drug absorption, reduced 51
dexamethasone 153, 160, 245, 296 drug binding, proteins involved in 125
dextromethorphan 218, 221 drug-excipient interactions 270
dextropropoxyphene 162 drug-food interactions 45
2,4-diaminopyrimidine 237 drug interactions
diamorphine 296 in drug analyses 265
diazepam 14,26, 100, 105, 106, 138, mechanisms 254
141,250,255,260,261,264,285, results of formulation changes 266
286,288,296 with packaging and intravenous
diazoxide 226,227,296 equipment 252
dichloralphenazone 296 drug-nutrient interactions 2
dichlorobiphenyl 164
diclofenac 65, 72, 182, 204, 296 E2020 101, 106
dicloxacillin 185 elderly
dicoumarol 14,162,296 absorption 284
didanosine 52, 57 distribution 284
dideoxycytidine 52 excretion 288
dietary fibres 332, 333 metabolism 287
diethylhexylphthalate(DEHP) 263, pharmacodynamics 289
264,296 pharmacokinetics 284
diethylpropion 272, 296 emodin 312
difiunisal 187 encainide 81
digitoxin 14, 141, 288 enfiurane 296
digoxin 15, 21, 22, 24, 37, 127, 136, 138, enoxacin 15,27,162,338
145,198,199,200,201,217,250, enprofylline 178
267,283,285,289,307,308,327, enzyme
333,334 induction 158
dihydrostreptomycin 286 inhibition 158, 160
diltiazem 65, 73, 81, 200, 217 secretion 47
assay 319 ephedrine 327
dimeclocycline 19 epilepsy, combination treatment 238
Subject Index 357

ergot 296 fluvoxamine 101, 106, 221, 222

ergotamine 217,222 frusemide 178, 282, 292, 296
ergot alkaloids 327 furazolidine 299,312
erythromycin 36,37,153,162,163,226, furosemide 310
296 fusidate sodium 66
acistrate 65, 73 fusidic acid 296
lactobionate 250
stearate 73 GABA 239
estrogen 36 GABAergic synapse 223
ethambutol 239 gabapentin 224
ethanol 15,26,153,224,227,228,288 garlic 340
ethchlorvynol 296 gastric emptying 48
ether, diethyl 296 residence time 47
ethinamate 296 gastrointestinal motility 22
ethinylestradiol 34, 100, 105 pH 22,26
17-ethinylestradiol 153 gastrointestinal tract, influence of food
ethosuximide 239, 240, 296, 297 on 46
ethotoin 296 genetic polymorphism 4
ethylmalonic acid 316 gentamicin 285, 286, 296, 344
ethylvinyl acetate 253 gentisic acid 309
etilefrine 217 gepirone 82
etomidate 296 ginseng 342
etoposide 263 glass, general properties 253
eucalyptol 296, 336 glibenclamide 162,163,226,334
eucalyptus species 336 glipizide 296, 306
exthoxazena 312 glomerular filtration 128, 173, 174, 289
rate 174,182,199,205
fadrozole 65 glucagon 160
famciclovir 68 glucocorticoids topical, raised
famotidine 66, 184, 195, 196,240 intraocular pressure and glaucoma
felbamate 162, 163 with 300
felodipine 35, 36, 49, 81, 82, 89, 332, D-glucose 29
336 glucose determination 306
fenoterol 34 glucose-6-phosphate dehydrogenase,
fenretidine 82, 89, 90 deficiency 297
fentanyl 296 glutethimide 161, 296
ferrous ion 15 glycaemic regulation 227
salts 312,313 glyceryl trinitrate 259, 260, 261, 264
sulfate 15, 16, 17 glycylglycine 31
fish oil 329 glycyrrhizin 336, 337, 343
FK 506 255 gold preparations 296
flecainide 53, 59, 141, 156 Gorei-san 344
fleroxacin 182 grapefruit juice 34, 36, 49, 89, 90, 332,
flucloxacillin 185 336
fluconazole 82, 162, 163 griseofulvin 246, 296
flucytosine 317,318 guanethidine 216, 218, 296
flufenamic acid 130, 296 guar gum 334
flumazenil 224 guvacoline 340
flunitrazepam 288, 296
fluoroquinolones 163 halofantrine 225,226
5-fluorouracil 246, 251, 255, 292 haloperidol 220, 226, 335
fluoxetine 221, 222 halothane 296
flupenthixol 220 heparin 138,139,282,296,313
fluphenazine 335 herbal drugs 330, 332
flurbiprofen 66, 73, 187,286,296 interactions with orthodox drugs
fluroxine 296 339
fluvastatin 66, 73 remedies 327
358 Subject Index

herbs, caffeine containing 338 kaolin 16

sparteine containing 338 kaolin-pectin 14
homatropine 15, 223 karela 341
homeostatic regulations 228 ketamine 296
hormonal systems 226 ketanserine 226
housekeeper wave 46 ketoconazole 34, 153, 162, 163
hydralazine 49, 53, 60, 296, 299, 330 ketoprofen 133,187,296
hydrochlorothiazide 15, 98, 99, 203, kidney, anatomy and physiology 173
226,296 n-kytorphin 29
hydroflumethiazide 202
hydroxychloroquine 66, 74 labetalol 296
7-hydroxy coumarin 332 laboratory tests, drugs causing
hyoscine 296 interference with 305
hyoscine-n-butylbromide 296 lactulose 310
Hyoscyamus niger 340 lamotrigine 135, 239
p-hydroxytriamterene 192,195 leaching 262
lead 327
ibuprofen 101,106,187,204,296 lecithins 222
idarubicin 312 L-Ieucine 29
imbrilon 273 levamisole 292
imidazolines 218 levodopa 15,26,30,53,60,82,91,101,
imipramine 28,141,216,217,218,223, 106,219,220,284,308,309,310,
226,286,288,296,298,335 312,329,331
immunoassays, interference with 319 levomepromazine 286
indigotin disulfonate 312 lidocaine 153,225,288, 318
indocid 273,274 lignocaine 15, 36, 141,246, 285, 296
indocyanine green 313 lincomycin 15
indomethacin 131,179,186,187,204, liquorice 337, 341
272,273,282,286,296,312,313 lithium 15,201, 202, 203, 204, 205, 286,
indoprofen 286 296,333
influenza vaccine 162 salts 333
insulin 6, 217, 227, 228, 243, 258, 259, lomefloxacin 66, 162
296 loop diuretics 203
induced hypoglycaemia 227 loracarbef 67, 74
interferon 158 lorazepam 138,286,288,296
intestinal metabolism 33 lorcainide 142
motility 47 loxapine 335
ion channels 224, 225
ipratopium 223 magnesium 262
iproniazid 160, 162 magnesium hydoxide 14,17,26,240
iron 14, 282, 283 stearate 272
iron sorbitex 312 sulfate 17
isoniazid 8, 15, 160,239,288,296,299, mandelamine 296
330 Mandragora officinarum 340
isoprenaline 290 mecamylamine 296
isopropyl antipyrine 297 meclozine 296
isopropyl meprobamate 296 mefenamic acid 296
isosorbide dinitrate 262 mefloquine 225, 226
isosorbide-5-mononitrate 66 melphalan 245, 286
ispaghula husk 333 meperidine 218, 288
itraconazole 35, 82, 91, 162, 163 mephenazine 296
mephenytoin 153, 296
Jaffe's reaction 310 meprobamate 296
mercaptopurine 162
kampo 336 mercury compounds 296
kanamycin 262,286,316 mersalyl 296
sulphate 250 metaboliser phenotype 156
Subject Index 359

metal ions, complexation with 18 receptor blocking drugs 223

metformin 53,188,192,243,296,298, mustine 245
334 myocardial infarction 126
methacrylate butadiene styrene 253
methacycline 15 nabumetone 95, 96, 97
methadone 157, 296 nadolol 16,21,285
methchlorethamine 130 NADPH 165
methenamine mandalate 296 nafcillin 181
methicillin 317 nalidixic acid 296
methocarbamol 312 naphthalene 299
methotrexate 53, 60, 67, 74, 101, 133, naphthylamine 153
184,185,186,187,255,287,316,318 naproxen 53,95,96,187,204,296
methoxsalen 341 naringin 332
methoxyflurane 296 navelbine 53
methoxypsoralen 83 nefiracetam 102
methsuximide 296, 297 neoarsphenamine 299
f3-methyldigoxin 64, 70, 71 neomycin 15, 16, 17
methyldopa 296,310,312 neostigmine 296
a-methyldopa 29, 216, 218 Nerium oleander 340
3-0-methyldopa 60 netilmicin 287
methylene blue 312 nialamide 160
3,4-methylenedioxymethamphetamine nicardipine 75, 163
220,221,226 nicorandil 67
methylmalonic acid 316 nicotine 188, 193, 195
methylphenidate 296 transdermal delivery systems 223
methylprednisolone 245 nicotinic acid 227
methyprylone 296 receptors 222
metoclopramide 14, 15,22, 28, 71, 142, nicoumalone 162
220,223,296 nifedipine 49, 67, 75, 81, 83, 90, 91, 153,
metoprolol 49, 55, 106, 107, 287, 288, 157,163,200,332,336
298 nikethamide 296
succinate 101 ninhydrin 316
metronidazole 242, 296, 312, 317 nitrates 259
metyrapone 296 nitrazepam 287, 288, 296
mexiletine 225, 298 nitrendipine 53, 60, 75, 332, 336
micelle formation 29 nitrofural 299
miconazole 256, 263 nitrofurantoin 16,296,299, 313, 318
midazolam 35, 153 nitroglycerin 260
migrating motor complex 46 nitrosamine 153
minerals 331 N-nitrosodiethylamine 153
misoprostol 240 nitrous oxide 296
mitomycin 255 nizatidine 240
mitoxantrone 312 non-steroidal anti-inflammatory drugs
MK-679 53, 60 186,203,217,228,229,230,290,
moclobemide 3, 83, 218, 220, 221, 222 292,313,327
monofluorophosphate 67, 74 noradrenaline 3, 216
moricizine 67, 162, 163 transporter 218
morphine 16, 34, 107, 152, 157, 161, noradrenergic synapse 215
274,287,296,327 norethindrone 100, 105
hydrochloride 274 norfloxacin 16,27,53,61,162
sulfate 101 nortriptyline 156, 163,246,287,288,
mosapride citrate 101 296,298,299
moxalactam 182, 183 novobiocin 296
moxonidine 101, 102, 107, 108 Nuelin 268
mucosal changes, drug induced 37
multidrug resistance gene 32 octreotide 227
muscarinic receptors 222 oestrogens 296
360 Subject Index

ofloxacin 15, 67, 162,336 phenformin 243, 244, 296

oleic acid 28 phenindione 313
omeprazole 57,153,240,241,242 pheniprazine 161
oncovin 37 pheniramine 296
orciprenaline 34 phenobarbital 287
organic cations, transport processes for phenobarbitone 16, 21, 153, 159, 239,
175 240,249,298
organic neutral drugs 198 phenol red 28
oxazepam 138,157,287,288,296 phenolphthalein 313
oxcarbazepine 83, 91 phenolsulphophthalein 313
oxidative genetic polymorphism 298 phenoperidine 296
oxpentifylline 296 phenothiazines 227,313
oxybutinin 83 phenoxybenzamine 297
oxytetracycline 15 phenoxymethylpenicillin 334
phenprocoumon 161
P-glycoprotein 32 phensuximide 296,297,313
paclitaxel 263 phenylalanine 316
pafenolol 33 phenylbutazone 8, 16, 21, 129, 130, 131,
pamaquine 299, 313 132,142,160,161,205,282,283,
pancuronium 296 287,288,297,298,307
Papaver somniferum 340 phenylephrine 217
paracellular pathway 29 phenylethylamine 3
paracetamol 16,53,61,68,72,75,153 phenylhydrazine 297,299
paraldehyde 296 phenylpropanolamine 218
paramethadione 296 phenylpyruvic acid 316
pargyline 296 phenytoin 16,21,53,83,92,133,135,
paroxetine 102, 152, 221 139, 142, 157, 159, 160, 161, 163,
penbutolol 130 227,239,240,246,251,267,283,
penciclovir 68 287,288,296,298,313,342
penicillamine 16, 296, 330 sodium 21
pencillin 176, 185, 236 physicochemical interactions 13
G 287 Picrorhiza kurroa 341
V 16 picrotoxin 224
penicillins 181, 185 Pilocarpus pennatifolius 340
pentagastrin 27 pindolol 188, 192, 193
pentamidine 227 pipemidic acid 338
pentaquine 299 piperacillin 185
pentazocine 159, 296 piperine 342
pentobarbitone 249 pirenzipine 223
pentylenetetrazol 29 piroxicam 205
peptic ulcer therapy 240 piroximone 102
perfloxacin 162 plasma binding displacement,
permeation through silicone therapeutic consequences 130
rubber 264 plasma protein binding, alteration in
through plastics 264 disease states 139
perphenazine 156 plasma pseudocholinesterase,
pethidine 16, 161, 163, 218, 221, 296, abnormalities in 299
297 plasma and tissue binding, influence on
pharmacodynamic interactions 215 drug kinetics 127
mechanisms 215 plastics, general properties 252
phase I biotransformations 34 sorption to 257, 258
phase II conjugations 34 polycyclic hydrocarbons 160
phenacemide 310 polyethylene 253
phenacetin 152, 153, 154, 313 polyethylene glycol 271
phenazone 159,297 polymer modification 264
phenazopyridine 313 polyoxyethylated castor oil 263
phenelzine 3,297,299 polypropylene 253
Subject Index 361

polysorbate 80, 263 pyrazinamide 239, 297

polystyrene 253 pyrazolone, sodium phenyl
polyvinyl alcohol 268 dimethyl 297
copolymer 270 pyridoxine 219,297, 329
methylacrylate 268, 269 hydrochloride 272
polyvinylchloride 253,257,260,262, pyrimethamine 237,297
263, 264, 265 pyrvinium pamoate 313
polyvinylpyrrolidone 271
potassium channels and drug-induced quinacrine 313
Torsades de Pointes 225 quinidine 136, 138, 139, 145, 153, 188,
ATP sensitive 226 194,195,199,223,225,226,246,
potassium homeostasis 229 287,288,292,299,307,308,316,
potassium-sparing diuretics 198 327,338,341
practolol 287 quinine 16,26,28,195,227,313,327,
pravastatin 53, 62 332
prazosin 217,288,298 quinocide 299
prednisolone 142,245,282,290,296,
327,336,337,338,344 ranitidine 15,27,100,105,188,191,
prednisone 37,272 194,195,240,242,297
prilocaine 296 Rauwolfia serpentina 340
primaquine 297,299,313 rectal bioavailability, formulation effects
primidone 160, 240, 297 on 272
probenecid 176,177,181,182,184,185, remikiren 62
236,297 renal excretory mechanisms, drug
procainamide 28, 102, 176, 188, 189, interactions involving 173
190, 191, 194, 195, 196, 197, 299, repirinast 83, 93
316,317 reserpine 216,218,219,221,297,327,
procaine 296 343
penicillin 287 resorcinol 297, 313
procarbazine 37, 245 riboflavin 313, 333
procyclidine 340 rifabutin 68, 102
progabide 224 rifampicin 8,17,21,34,160,161,239,
progesterone 83 240,297,342
progestogens 297 rifampin 313
proguanil 163 riboflavine-5f-phosphate 16
proguazone 16 R042-5892 53
promazine 16 rufloxacin 53
promethazine 223, 250, 296, 313 rutin 343
propafenone 49, 156, 161
propanidid 296,297 S-1l08 83,93
propantheline 14,15,22,25,71 Sairei-to 344
bromide 296 salazosulphapyridine 299
propicillin 287 salicylate 131,133,142,157,171,205,
propofol 313 206,309,313
propoxyphene 159,297 salicylic acid 130,317, 331
hydrochloride 271 salsalate 68, 76
propranolol 49, 55, 100, 105, 106, 126, santonin 313
128,139,142,282,287,288,297, Scopalia carniolica 340
298,342 selegiline 162, 163
prostaglandin synthesis 206 senna 313
prostaglandins 228, 229 sennosides 327
prostigmine 297 serotonergic synapse 200, 221
proteolytic enzymes 17 serotonin 3,220,221,222
prothrombin time 131 sertraline 221
protriptyline 287 shankhapusphi 342
pseudoephedrine 16, 79, 102, 206 Sho-saiko-to 344
psyllium 333 signal transduction 155
362 Subject Index

SK&F 106203 54, 62 terfenadine 35, 68, 163, 226

Slo-Phyllin 268, 269 tetracycline 3, 15, 17, 18, 37, 54, 61,
small peptide transporter . 31 242,250,283,287,331,342
sodium alginate 16 tetraethylammonium bromide 297
sodium bicarbonate 17,202 Teucrium chamaedrys 339
sodium cholate 16 Theo-24 93, 268
sodium lauryl sulfate 28 Theobid 269
sodium taurocholic acid 29 Theo-Dur 268,269,270
sodium taurodeoxycholic acid 28 Theophyl SR 269
somatostatin 226 theophylline 68, 69, 76, 84, 93, 103, 109,
sotalol 54, 62, 226 143,160,161,163,201,202,246,
sparfloxacin 83, 102, 108 268,269,287,288,289,297,317,
sparteine 153, 156, 344 318,342
spironolactone 198,282,287,297 therapeutic drug monitoring 134
splanchnic blood flow 47 thiazide diuretics 202, 227
Stevens-Johnson syndrome 238 thiazosulphone 299,313
stomach emptying rate 47 thiopentone 152, 250
streptomycin 297 thioridazine 287,298
streptozotocin 227 thyroid hormone 227
succinyl choline 297 thyroxine 17,20,283,331
sulfadiazine 17, 26, 243, 342 tiagabine 69, 76
sodium 17, 250 tiaprofenic acid 103, 108, 109
sulfadimethoxine 28 ticarcillin 182, 185
sulfaguanidine 28 ticlopidine 84, 93
sulfamethizole 28, 287 tight junctions 29
sulfamethoxazole 17,133,196,237 tissue binding displacement, therapeutic
sulfanilacetamide 28 consequences 136
sulfanilamide 28, 271, 299 tobacco smoke 153,160
sulfapyridine 26, 299 tobramycin 287
sulfathiazole 17 tolazamide 297
sulfisoxazole 28, 143 tolbutamide 126, 133, 153, 157, 159,
sulindac 187, 205, 206, 228 226,243,283,287,297
sui penicillin 287 tolfenamic acid 143
sulphacetamide 299 tolonium 313
sulphamethoxypyridazine 299 tolrestat 244
sulphasalazine 265, 313 toluidine blue 299
sulphinpyrazine 133, 179 topiramate 69
sulphonal 297 TPN fluid 251, 252, 263, 264
sulphonylurea induced hypoglycaemia tramadol 84
227 transmitter systems 215
sulpiride 54, 62, 219 tranylcypromine 3,297
sulthiame 297 trazodone 69
sumatriptan 222 triacetyloleandomycin 162
sustained-release formulation, problems triamterine 188, 191, 192, 195, 313
with 268 trichloracetic acid 130
suxamethonium 299, 300 trifluoperazine 297, 335
trihexylphenidyl 26, 223
tacrine 54, 62 trimetazidine 103
tacrolimus 163 trimethadione 296
talinolol 33 trimethoprim 17,133,196,237,238
tamoxifen 162,163 trimethylamine 152, 165
tannins 334 trinitrotoluene 299
taxotene 263 trional 297
temafloxacin 103, 108 tripelennamine 297
temazepam 288 trometamol 251
teniposide 263 troxidone 297
terazocin 68 L-tryptophan 329
Subject Index 363

tubocurarine 297,327 vincnstme 32, 245

tubular reabsorption 177, 201 vinpocetine 84, 94
tubular secretion 184 vitamin A 17,256,334
drug interactions involving 177 vitamin B6 330
evidence for 176 vitamin B12 17, 330
tyramine 3, 98, 164, 165,216,218,219, vitamin C 297,330
220 vitamin D 256, 330
vitamin E 256, 334
Uniphyllin 268 vitamin K 330, 331
University Group Diabetes Program vitamin K4 330
(UGDP) 243
Urginea maritima 340 warfarin 8,17,20,128,129,130,131,
uric acid 306, 309 133,139,143,153,159,163,246,
urinary alkalinizers 185 256,260,282,283,285,287,288,
urine amino acid screening 315 289,292,307,313,331,340,341
urine PI:I and flow rate, effect on drug S-warfarin 132
clearance 206 World Health Organization 293
urine protein reagent strip test 314
xanthine oxidase 165
valproate sodium 76, 95, 96, 133, 135, xylocaine 297
239, 240, 297 Xysmalobium undulatum 340
valproic acid 69, 78, 96, 130, 140, 143,
297 yohimbine 343
vanoxerine 84, 94
ventricular tachycardia 35 zalospirone 69, 84, 94
verapamil 32, 54, 62, 103, 109, 110, 126, zidovudine 54,62,69,77,180,196,
134,143,161,200,217,282 197
vigabatrine 69, 224, 239, 240 zimeldine 221
viloxazine 297 zinc 331
vinblastine 32, 255 sulfate 17
vinca alkaloids 327 zomepirac 143
 Spri nger-Verlag
and the Environment

We at Springer-Verlag firmly believe that an

international science publisher has a special
obligation to the environment, and our corpo-
rate policies consistently reflect th is conviction.

We also expect our busi-

ness partners - paper mills, printers, packag-
ing manufacturers, etc. - to commit themselves
to using environmentally friendly materials and
production processes.

The paper in this book is made from

low- or no-chlorine pulp and is acid free, in
conformance with international standards for
paper permanency.